Datasets:

debajyotidasgupta

/

dblp_2022_2023

like 0

Title: Doubly Robust Crowdsourcing Abstract: Large-scale labeled dataset is the indispensable fuel that ignites the AI revolution as we see today. Most such datasets are constructed using crowdsourcing services such as Amazon Mechanical Turk which provides noisy labels from non-experts at a fair price. The sheer size of such datasets mandates that it is only feasible to collect a few labels per data point. We formulate the problem of test-time label aggregation as a statistical estimation problem of inferring the expected voting score. By imitating workers with supervised learners and using them in a doubly robust estimation framework, we prove that the variance of estimation can be substantially reduced, even if the learner is a poor approximation. Synthetic and real-world experiments show that by combining the doubly robust approach with adaptive worker/item selection rules, we often need much lower label cost to achieve nearly the same accuracy as in the ideal world where all workers label all data points.

Title: A Generalized Minimax Q-Learning Algorithm for Two-Player Zero-Sum Stochastic Games Abstract: We consider the problem of two-player zero-sum games. This problem is formulated as a min–max Markov game in this article. The solution of this game, which is the min–max payoff, starting from a given state is called the min–max value of the state. In this article, we compute the solution of the two-player zero-sum game, utilizing the technique of successive relaxation that has been successfully applied in this article to compute a faster value iteration algorithm in the context of Markov decision processes. We extend the concept of successive relaxation to the setting of two-player zero-sum games. We show that, under a special structure on the game, this technique facilitates faster computation of the min–max value of the states. We then derive a generalized minimax Q-learning algorithm, which computes the optimal policy when the model information is not known. Finally, we prove the convergence of the proposed generalized minimax Q-learning algorithm utilizing stochastic approximation techniques, under an assumption on the boundedness of iterates. Through experiments, we demonstrate the

Title: A robot’s sense-making of fallacies and rhetorical tropes. Creating ontologies of what humans try to say Abstract: In the design of user-friendly robots, human communication should be understood by the system beyond mere logics and literal meaning. Robot communication-design has long ignored the importance of communication and politeness rules that are ‘forgiving’ and ‘suspending disbelief’ and cannot handle the basically metaphorical way humans design their utterances. Through analysis of the psychological causes of illogical and non-literal statements, signal detection, fundamental attribution errors, and anthropomorphism, we developed a fail-safe protocol for fallacies and tropes that makes use of Frege’s distinction between reference and sense, Beth’s tableau analytics, Grice’s maxim of quality, and epistemic considerations to have the robot politely make sense of a user’s sometimes unintelligible demands.

Title: What Machine Learning Can Learn From Software Modularity Abstract: More program functions are no longer written in code but learned from a huge number of data samples using a machine learning (ML) algorithm. We present an overview of current techniques to manage complex software and discuss how this applies to ML models.

Title: Coherence Constrained Graph LSTM for Group Activity Recognition Abstract: This work aims to address the group activity recognition problem by exploring human motion characteristics. Traditional methods hold that the motions of all persons contribute equally to the group activity, which suppresses the contributions of some relevant motions to the whole activity while overstating some irrelevant motions. To address this problem, we present a Spatio-Temporal Context Cohere...

Title: Pose-Guided Representation Learning for Person Re-Identification Abstract: The large pose variations and misalignment errors exhibited by person images significantly increase the difficulty of person Re-Identification (ReID). Existing works commonly apply extra operations like pose estimation, part segmentation, etc., to alleviate those issues and improve the robustness of pedestrian representations. While boosting the ReID accuracy, those operations introduce considerab...

Title: Hierarchical Deep Click Feature Prediction for Fine-Grained Image Recognition Abstract: The click feature of an image, defined as the user click frequency vector of the image on a predefined word vocabulary, is known to effectively reduce the semantic gap for fine-grained image recognition. Unfortunately, user click frequency data are usually absent in practice. It remains challenging to predict the click feature from the visual feature, because the user click frequency vector of an ...

Title: P-CNN: Part-Based Convolutional Neural Networks for Fine-Grained Visual Categorization Abstract: This paper proposes an end-to-end fine-grained visual categorization system, termed Part-based Convolutional Neural Network (P-CNN), which consists of three modules. The first module is a Squeeze-and-Excitation (SE) block, which learns to recalibrate channel-wise feature responses by emphasizing informative channels and suppressing less useful ones. The second module is a Part Localization Network...

Title: Fine-Grained Video Captioning via Graph-based Multi-Granularity Interaction Learning Abstract: Learning to generate continuous linguistic descriptions for multi-subject interactive videos in great details has particular applications in team sports auto-narrative. In contrast to traditional video caption, this task is more challenging as it requires simultaneous modeling of fine-grained individual actions, uncovering of spatio-temporal dependency structures of frequent group interactions, an...

Title: Nearest neighbor imputation for categorical data by weighting of attributes Abstract: Missing values are a common phenomenon in modern medical research of complex diseases. The data often contains nominal or categorical variables, for example, single nucleotide polymorphisms (SNPs) in genetic studies. If the missing values are not handled properly, the downstream statistical analysis of incomplete data may be biased. While various imputation methods are available for metrically scaled variables, methods for categorical data are scarce. An imputation method that has been shown to work well for high dimensional metrically scaled variables is the imputation by nearest neighbor methods. In this paper, we propose a weighted nearest neighbors approach to impute missing values in categorical variables in high dimensional datasets. The proposed method explicitly uses the information on the association among attributes. Using different simulation settings, the performance is compared with available imputation methods. A variety of real data sets, containing heart, DNA, and lymphatic cancer, is also used to support the results obtained by simulations. The results show that the weighting of attributes yields smaller imputation errors than existing approaches like random forest and MICE.

Title: TROPICAL LINES ON CUBIC SURFACES Abstract: Given a tropical line L and a smooth tropical surface X, we look at the position of L on X. We introduce its primal and dual motifs which are respectively a decorated graph and a subcomplex of the dual triangulation of X. They encode the combinatorial position of L on X. We classify all possible motifs of tropical lines on general smooth tropical surfaces. This classification allows us to give an upper bound for the number of tropical lines on a general smooth tropical surface with a given subdivision. We focus in particular on surfaces of degree three. As a concrete example, we look at tropical cubic surfaces dual to a fixed honeycomb triangulation, showing that a general surface contains exactly 27 tropical lines.

Title: Performance validation of clustering algorithms using selection of attributes and application of filters in terms of data reduction Abstract: Clustering in unsupervised learning is a method to find inherent group of set of unlabeled data. Such set of groups are termed as Clusters. Grouping of datasets in to clusters involves minimization of the interclass similarity and maximization of the intraclass similarity. Therefore, clustering the large datasets introduces the concept called Data Reduction. Data reduction is a simple process of identifying a relevant feature subset, which is enough to represent the selected large datasets. Here, the data reduction is done in terms of applying two clustering algorithms like Expectation and Maximization-EM and K-Means with filters in unsupervised category (i) Normalize filter in Instance level and (ii) Randomize filter in attribute level and selection of attributes. The results shows that the Livestock dataset applied with selection of attributes and preprocessing filter named Normalize in attribute level and Randomize in instance level on which the EM and K-Means algorithm are executed, and the results are compared and analyzed in terms of data reduction. Therefore, the K-Means algorithm applied with Randomize filter at instance level and selection of attributes and finally K-Means algorithm show good performance in case of large datasets when compared with other clustering algorithms.

Title: Analysis of attribute-based cryptographic techniques and their application to protect cloud services Abstract: Recent technological advances such as the Internet of things, fog computing, and cloud applications lead to exponential growth in the amount of generated data. Indeed, cloud storage services have experienced unprecedented usage demand. The loss of user control over their cloud stored data introduced several security and privacy concerns. To address these concerns, cryptographic techniques are widely adopted at the user side. Attribute-based cryptography is commonly used to provide encrypted and/or authenticated access to outsourced data in remote servers. However, the use of these cryptographic mechanisms often increase the storage and computation costs; consequently, the energy consumption in the entire cloud ecosystem. In this paper, we provide a comparative analysis of different attribute-based cryptographic mechanisms suitable for cloud data sharing services. We also provide a detailed discussion of different reviewed schemes, w.r.t. supported features, namely, security, privacy, and functional requirements. in addition, we explore the limitations of existing attribute-based cryptographic mechanisms and propose future research directions to better fit the growing needs of this cloud environment in terms of energy savings, processing and storage efficiency, and availability requirements.

Title: BlackEye: automatic IP blacklisting using machine learning from security logs Abstract: Blacklisting of malicious IP address is a primary technique commonly used for safeguarding mission-critical IT systems. The decision to blacklist an IP address requires careful examination of various aspects of packet traffic data as well as the behavioral history. Most of the current security monitoring for IP blacklisting heavily relies on the domain expertise from experienced specialists. Although there are efforts to apply machine-learning (ML) techniques to this problem, we are yet to see the mature solution. To mitigate these challenges and to gain better understanding of the problem, we have designed the BlackEye framework in which we can apply various ML techniques and produce models for accurate blacklisting. From our analysis results, we learn that multi-staged method that combines the data cleansing and the classification via logistic regression or random forest produces the best results. Our evaluation on the real-world data shows that it can reduce the the incorrect blacklisting by nearly 90% when compared to the performance of experts. More over, our proposed model performed well in terms of the time-to-blacklist by curtailing the period of malicious IP address in activity by 27 days on average.

Title: Double-Blinded Finder: a two-side secure children face recognition system Abstract: When taking photos of a suspicious missing child in the street and posting them to the social network is becoming a feasible way to find missing children, the exposure of photos may cause privacy issues. To address this problem, we propose Double-Blinded Finder, an efficient and double-blinded system for finding missing children via low-dimensional multi-attribute representation of child face and blind face matching. To obtain enough knowledge for representing child faces, we build the Labeled Child Face in the Wild dataset, which contains 60K Internet images with 6K unique identities. Based on this dataset, we further train a multi-task deep face model to describe a child face as a 128d fixed-point feature vector and extensive gender and age attributes. Using the keys generated from face descriptors, the face photos from the social network and face representation from the parent(s) of missing children are both encrypted. In addition, we devise inner-production encryption to run blind face matching in the public cloud. In this manner, Double-Blinded Finder can provide efficient face matching while protecting the privacies of both sides: (1) the suspicious missing children side for avoiding the invasion of the human rights, and (2) the true missing children side for preserving the secondary victimization. The experiments show that our system can achieve practical performance of child face matching with negligible leakage of privacy.

Title: Securing password using dynamic password policy generator algorithm Abstract: It is proposed to tackle the problem of password leakage of popular websites like Linked-In, Adobe, Gmail, Yahoo, eHarmony, etc. by using dynamic password policy and enhanced hash algorithm. Here, an algorithm is developed that will generate password policies dynamically depending on the frequency of characters. Time complexity is computed, and it is found that the algorithm works fast. Since the algorithm creates password policies dynamically, it will be tough for the attacker to guess the characteristics of the password database.

Title: Exploring online peer feedback and automated corrective feedback on EFL writing performance Abstract: Previous studies have been done to research the effects of different electronic feedback (e-feedback) modes of helping English as a foreign language (EFL) students improve their writing. The purpose of this study was to employ online peer feedback (OPF) and automated corrective feedback (ACF) to assess EFL learners' writing performance in the areas of sentence complexity, grammatical accuracy, and lexical density. Our major findings suggest that OPF is potentially more useful than ACF in improving sentence writing, making fewer grammatical errors, and producing more types of lexical items. Nevertheless, we also found that the majority of students favored the use of ACF for the purpose of producing a richer vocabulary. Less skilled writers tended to make greater improvements than higher-skilled writers in producing more sentences and lexical items after OPF use; however, the difference was not statistically significant. Based on our research results, we discuss and present the implications of these findings for pedagogical instruction and future research.

Title: Neural network-based leaf classification using machine learning Abstract: In this paper, the leaf image data set is taken for classification. The process is done as image pre-processing, image segmentation, feature extraction, and image classification. The segmentation is done on colours. The K-Means clustering is applied to group the similar colour pixels. In total, three sub-images are created as an output of segmentation. In total, three classes are considered for the process. Data set with 70 leaf images is considered for the classification process. The data set are classified as yellow-based leaves, brown-based leaves, and green-dominated leaves. In this, the first two classes are considered as infected, based on the colours, and the third class is an uninfected collection of leaves. Seven different Neural Networks are constructed for classification process using MATLAB. Their performances are evaluated using the confusion matrices. From the outcome of confusion matrix, it is clear that the Regression Neural Network and Radial Bias Neural Network are better classifiers out of the seven architectures.

Title: A dynamic wrapper-based feature selection for improved precision in content-based image retrieval Abstract: The content-based image retrieval (CBIR) has been applied in the image processing as well as pattern recognition. A challenging task in the CBIR research has been through the feature extraction for decreasing any semantic gap that is an active research topic. Here, in this work, there is a texture feature that is extracted from an image, making use of a technique of curvelet transform. This curvelet is selected for a sparse representation that is quite critical for the estimation of images, which have been de-noised with some inversed problems. The wrapping-based curvelet transform will be even more robust as well as faster in the time of computation than the Ridge Transform. A technique of feature selection will be brought for selecting the optimal features. The correlation-based feature selection (CFS) method has been adapted for improving the accuracy of the CBIR systems. The non-deterministic polynomial (NP)-hard problems may be solved using the chemical reaction optimization (CRO) that has been motivated using the technique of chemical reaction. The benefits of proposed method is a necessity for marginal human intrusion for the purpose of retrieving the images needed from its database. The proposed mechanism has been again assessed depending on the Coral Database, and a performance study is ended using precision, f measure, and recall.

Title: Benchmarking on offline Handwritten Tamil Character Recognition using convolutional neural networks Abstract: Convolutional Neural Networks (CNN) are playing a vital role nowadays in every aspect of computer vision applications. In this paper we have used the state of the art CNN in recognizing handwritten Tamil characters in offline mode. CNNs differ from traditional approach of Handwritten Tamil Character Recognition (HTCR) in extracting the features automatically. We have used an isolated handwritten Tamil character dataset developed by HP Labs India. We have developed a CNN model from scratch by training the model with the Tamil characters in offline mode and have achieved good recognition results on both the training and testing datasets. This work is an attempt to set a benchmark for offline HTCR using deep learning techniques. This work have produced a training accuracy of 95.16% which is far better compared to the traditional approaches.

Title: Clustering-based real-time anomaly detection-A breakthrough in big data technologies Abstract: Off late, the ever increasing usage of a connected Internet-of-Things devices has consequently augmented the volume of real-time network data with high velocity. At the same time, threats on networks become inevitable; hence, identifying anomalies in real time network data has become crucial. To date, most of the existing anomaly detection approaches focus mainly on machine learning techniques for batch processing. Meanwhile, detection approaches which focus on the real-time analytics somehow deficient in its detection accuracy while consuming higher memory and longer execution time. As such, this paper proposes a novel framework which focuses on real-time anomaly detection based on big data technologies. In addition, this paper has also developed streaming sliding window local outlier factor coreset clustering algorithms (SSWLOFCC), which was then implemented into the framework. The proposed framework that comprises BroIDS, Flume, Kafka, Spark streaming, SparkMLlib, Matplot and HBase was evaluated to substantiate its efficacy, particularly in terms of accuracy, memory consumption, and execution time. The evaluation is done by performing critical comparative analysis using existing approaches, such as K-means, hierarchical density-based spatial clustering of applications with noise (HDBSCAN), isolation forest, spectral clustering and agglomerative clustering. Moreover, Adjusted Rand Index and memory profiler package were used for the evaluation of the proposed framework against the existing approaches. The outcome of the evaluation has substantially proven the efficacy of the proposed framework with a much higher accuracy rate of 96.51% when compared to other algorithms. Besides, the proposed framework also outperformed the existing algorithms in terms of lesser memory consumption and execution time. Ultimately the proposed solution enable analysts to precisely track and detect anomalies in real time.

Title: A constrained framework for context-aware remote E-healthcare (CARE) services Abstract: Recently, due to the health awareness among the current generations, it leads us to give special attention to the remote E-healthcare services. The remote E-healthcare services have been categorized mainly into three categories: primary care, secondary care, and emergency care. Due to the categorized remote healthcare services, the resources must have to be utilized efficiently. Therefore, context-aware modelling for the remote E-healthcare (CARE) services has been used to save the resource wastage. The CARE services have been divided into three categories primary context-aware remote E-healthcare, secondary context-aware remote E-healthcare, and emergency context-aware remote E-healthcare (E-CARE). Three different algorithms have been proposed for the performance analysis. The result shows that the proposed algorithms for the primary context-aware remote E-healthcare and secondary context-aware remote E-healthcare do not make more difference in the routing. However, the selection of the path for the E-CARE shows that the hop count of the path has been increased. The increase in the hop count leads to an increase in energy efficiency and the life of the services. Additionally, observation made from our assigned algorithm E-CARE shows that it performs extremely well for the E-healthcare services without incrementing the time complexity. Furthermore, major characteristics of the algorithm has been explored and analyzed for the E-healthcare services.

Title: A Quality-of-Things model for assessing the Internet-of-Things' nonfunctional properties Abstract: The Internet of Things (IoT) is in a "desperate" need for a practical model that would help in differentiating things according to their nonfunctional properties. Unfortunately, despite IoT growth, such properties either lack or ill-defined resulting into ad hoc ways of selecting similar functional things. This paper discusses how things' nonfunctional properties are combined into a Quality-of-Things (QoT) model. This model includes properties that define the performance of things' duties related to sensing, actuating, and communicating. Since the values of QoT properties might not always be available or confirmed, providers of things can tentatively define these values and submit them to an Independent Regulatory Authority (IRA) whose role is to ensure fair competition among all providers. The IRA assesses the values of nonfunctional properties of things prior to recommending those that could satisfy users' needs. To evaluate the technical doability of the QoT model, a set of comprehensive experiments are conducted using real data sets. The results depict an acceptable level of the QoT estimation accuracy.

Title: Adaptive neuro-fuzzy inference system-based energy conservation system for performance enhancement of MANET Abstract: A mobile ad hoc network (MANET) is a network with dynamic topology and mobile nodes. Due to its dynamic nature, there is no central control. Hence, nodes communicate with other nodes through intermediate nodes. The intermediate nodes are normal nodes in the same network and assume the responsibility of forwarding packets on the route from source to destination. Along these lines, a MANET requires vitality effective directing conventions to sort out the power use and to improve battery usage. In this paper, an equipped convention is recommended that expends less vitality of portable hubs in MANETs that depend on the remaining edge vitality of adaptive network-based fuzzy inference systems (ANFIS). The edge esteem is resolved dependent on the fluffy participation capacities. Basically, the area hubs of source are figured utilizing two components: encased locale and transferring district. The neighbor hubs pondered by those elements receive the greatest edge esteem that started utilizing ANFIS. The hub with the most extreme estimations of edge and separation is favored as the following bounce for the exchange. The exploratory outcome delineates that our convention can recognizably diminish the aggregate vitality utilization for all effectively transmitted bundles and expand the lifetimes of hubs, overwhelmingly in the high versatility condition.

Title: The GOP Inter Prediction of H.264 AV\C Abstract: Video coding process is consisting of two parts of (encoding and decoding) the digital video scenes. The digital video is a illustration or representation of many scenes that told different stories. It captured from digital video camera and stored in a memory, so we can send video to another person through the internet. The translation of this type of videos needs a high bandwidth, high speed and huge amount of memory due to a huge amount of data to be translate. In the other side, the person needs to receive the video in a high accuracy, so we need a new way to achieve these options. H.264 advance video coding is a new technical leap to reduce the video size through the translation by compressing the original video and decompress the translated video at the receiver side to gain the same original video. The aim of the project is to achieve all these properties by building the encoder and decoder blocks of H.264 AV/C by using MATLAB. (C) 2019 The Authors. Production and hosting by Elsevier B.V. on behalf of King Saud University.

Title: Periocular biometrics: A survey Abstract: Periocular region is the feature rich region around the eye which may include features like eyelids, eyelashes, eyebrows, tear duct, eye shape, skin texture and many more. Periocular region based authentication system is a good trade-off between face and iris based biometric authentication systems as they need high user cooperation. This paper provides a comprehensive survey of periocular biometrics and a deep insight of various aspects such as utility of periocular region as a stand-alone modality, periocular region and its fusion with iris, application of periocular region in smart phone authentication and the role of periocular region in soft biometric classification etc. The paper also provides an outlook over possible future research in the area of periocular biometrics.

Title: Document representation and classification with Twitter-based document embedding, adversarial domain-adaptation, and query expansion Abstract: Document vectorization with an appropriate encoding scheme is an essential component in various document processing tasks, including text document classification, retrieval, or generation. Training a dedicated document in a specific domain may require large enough data and sufficient resource. This motivates us to propose a novel document representation scheme with two main components. First, we train TD2V, a generic pre-trained document embedding for English documents from more than one million tweets in Twitter. Second, we propose a domain adaptation process with adversarial training to adapt TD2V to different domains. To classify a document, we use the rank list of its similar documents using query expansion techniques, either Average Query Expansion or Discriminative Query Expansion. Experiments on datasets from different online sources show that by using TD2V only, our method can classify documents with better accuracy than existing methods. By applying adversarial adaptation process, we can further boost and achieve the accuracy on BBC, BBCSport, Amazon4, 20NewsGroup datasets. We also evaluate our method on a specific domain of sensitivity classification and achieve the accuracy of higher than $$95\%$$ even with a short text fragment having 1024 characters on 5 datasets: Snowden, Mormon, Dyncorp, TM, and Enron.

Title: Palm tree structure based data aggregation and routing in underwater wireless acoustic sensor networks: Agent oriented approach Abstract: The primary constraints of Underwater Wireless Acoustic Sensor Networks (UWASNs) are limited bandwidth, node energy and latency. The process of data aggregation will ease the constraints of UWASN. In this paper, we propose a scheme for data aggregation and routing in UWASN using static and mobile agents based on palm tree structure. The main components of palm tree structure include leaflets, spines, rachis and petioles. The leaflets and smaller leaflets (fronds) are connected to the petiole through spines. The junction of petioles connects to the sink node. The proposed scheme works as follows.Firstly, fronds and leaflets connected to the petioles are created through spines. Secondly, master center nodes are selected on petiole junction using mobile agent based on the parameters such as residual energy, Euclidean distance, petiole angle and connectivity. Thirdly, local center nodes are identified on either side of leaflet and connected to the master center using a mobile agent. Fourthly, the process of local aggregation at local centers happens by taking into account of nodes along the leaflets and carry to a connected master center. Finally, the process of master aggregation at the junction of petioles and delivering the aggregated data to the sink node using a mobile agent is performed. To assess the efficacy of the scheme, simulation in different UWASN scenarios are carried out. The parameters of performance analyzed are master and local center selection time, aggregation ratio, aggregation energy, energy consumption, number of local and master centers involved in the aggregation process and lifespan of the network. We observed that proposed scheme performs better than the existing aggregation scheme.

Title: Fuzzy dragon deep belief neural network for activity recognition using hierarchical skeleton features Abstract: In computer vision, human activity recognition is an active research area for different contexts, such as human–computer interaction, healthcare, military applications, and security surveillance. Activity recognition is performed to recognize the goals and actions of one or more people from a sequence of observations based on the actions and the environmental conditions. Still, there are numbers of challenges and issues, which motivate the development of new activity recognition method to enhance the accuracy under more realistic conditions. This paper proposes an error-based fuzzy dragon deep belief network (error-based DDBN), which is the integration of fuzzy with DDBN classifier, to recognize the human activity from a complex and diverse scenario, for which the keyframes are generated based on the Bhattacharya coefficient from the set of frames of the given video. The key frames from the Bhattacharya are extracted using the scale invariant feature transform, color histogram of the spatio-temporal interest dominant points, and hierarchical skeleton. Finally, the features are fed to the classifier, where the classification is done using the proposed error-based fuzzy DDBN to recognize the person. The experimentation is performed using two datasets, namely KTH and Weizmann for analyzing the performance of the proposed classifier. The experimental results reveal that the proposed classifier performs the activity recognition in a better way by obtaining the maximum accuracy of 1, a sensitivity of 0.99, and the specificity of 0.991.

Title: The influence of prior knowledge on the effectiveness of guided experiment design Abstract: Inquiry learning is an effective learning approach if learners are properly guided. Its effectiveness depends on learners' prior knowledge, the domain, and their relationship. In a previous study we developed an Experiment Design Tool (EDT) guiding learners in designing experiments. The EDT significantly benefited low prior knowledge learners. For the current study the EDT was refined to also serve higher prior knowledge learners. Two versions were created; the "Constrained EDT" required learners to design minimally three experimental trials and apply CVS before they could conduct their experiment, and the "Open EDT" allowed learners to design as many trials as they wanted, and vary more than one variable. Three conditions were compared in terms of learning gains for learners having distinct levels of prior knowledge. Participants designed and conducted experiments within an online learning environment that (1) did not include an EDT, (2) included the Constrained EDT, or (3) included the Open EDT. Results indicated low prior knowledge learners to benefit most from the Constrained EDT (non-significant), low-intermediate prior knowledge learners from the Open EDT (significant), and high-intermediate prior knowledge learners from no EDT (non-significant). We advocate guidance to be configurable to serve learners with varying levels of prior knowledge.

Title: Effect on keyboard-based English word acquisition Abstract: Recent years, dramatic changes in the mode of writing have occurred. Computers and other digital devices are increasingly replacing writing by hand. The sensory-motor experiences of typing (e.g. visual, haptic, motor) are different from those used in handwriting. Therefore, the influence and effect of keyboarding on linguistic performance, compared to handwriting or letter tiles, has seen much inquiry. The results of these studies have been mixed. In this study, researchers used grounded theory to explore the experiences and perceptions of keyboard-based English word acquisition learners. The participants came from various ages and backgrounds (30 English training institute kids, 46 elementary school students, 7 college students, 14 in-service teachers), as well as 10 parents of lower age learners (N = 107). Audio of the interviews was recorded digitally, transcribed verbatim and imported into NVivo 11.0 for coding and analysis. Overall, the participants confirmed the effect of keyboarding on English word learning, and a majority expressed positive attitudes towards keyboarding. Participants also listed which aspects of keyboarding interested them, such as improved learning efficiency, engaging learning activities or other incidental benefits.

Title: A multi-attack resilient lightweight IoT authentication scheme Abstract: Majority of the devices in the Internet of Things (IoT) are deployed in an environment that is susceptible to cyber-attacks. Due to the resource-constraint nature of IoT, it is very hard to meet the security challenges that arise due to the deployment of IoT devices in the unsecure environment. In this context, the authentication of IoT devices is one of the core challenges. Many protocols have been designed to address and overcome the security issues that stem from the authentication failure. However, many of these protocols are designed using the complex cryptographic techniques that may not be supported by IoT devices. In this paper, we propose a lightweight and secure mutual authentication scheme for resource constraint IoT devices. The proposed scheme is robust against cyber-attacks, such as impersonation, modification, session key disclosure, and eavesdropping attacks. The security of the proposed scheme is formally tested using the Automated Validation of Internet Security Protocols and Applications tool and found the scheme to be secure in the Dolev-Yao attack model. Moreover, the performance features such as communication overhead, computation time, and the turnaround time of the proposed scheme are evaluated and compared with the recent schemes of same category, where the proposed scheme shows a balance of performance without compromising the security features.

Title: Tumbling through tertiary education: an investigation of the use of Tumblr within a child development course Abstract: E-learning emphasizes collaborative learning through the use of Web 2.0. The tools of Web 2.0 advocate open collaboration, interactive technology, and personalized learning that promotes expanding the scope of teacher/student/course interaction, as well as advancing the social construction of knowledge, an important component of socio-cultural theory. Using writing as inquiry, we investigated the usefulness of using Tumblr (a microblog) as a teaching tool for three cohorts of students across two countries. The multimedia format of Tumblr enhanced student interaction with the material and with each other, while increasing the digital competencies of our students and ourselves. Tumblr also allowed for a personalization of the self-guided learning, and fostered a dynamic online learning community.

Title: China's high-tech industry efficiency measurement with virtual frontier data envelopment analysis and Malmquist productivity index Abstract: Data envelopment analysis (DEA) is a widely used non-parametric method in efficiency measurement with multiinputs and multioutputs. Malmquist productivity measures the efficiency change in different periods and decomposes the general efficiency change into technical efficiency change and frontier shift. In this paper, we choose a driving industry in social development, the high-tech industry, as an example to illustrate a new method, virtual frontier DEA model, in the aspect of improvement of the traditional DEA model. Additionally, we decompose the Malmquist productivity index with virtual frontier DEA model to find out the driving force of high-tech efficiency change.

Title: E-Had: A distributed and collaborative detection framework for early detection of DDoS attacks Abstract: •A Hadoop based distributed detection framework called E-Had to detect DDoS attacks is proposed.•E-Had distribute computational and memory overheads to multiple mappers and reducers.•E-Had is robust as it can continue to work in case some of the mappers do not respond in time.•E-Had is implemented using HA-DDoS testbed consisting of 30 real systems.•E-Had has been validated using different attack scenarios of CAIDA and DDoSTB datasets.

Title: Learning anytime, anywhere: a spatio-temporal analysis for online learning Abstract: The study proposes two new measures, time and location entropy, to depict students' physical spatio-temporal contexts when engaged in an online course. As anytime, anywhere access has been touted as one of the most attractive features of online learning, the question remains as to the success of students when engaging in online courses through disparate locations and points-in-time. The procedures describe an analysis of 5293 students' spatio-temporal patterns using metadata relating to place and time of access. Grouping into segments that describe their patterns of engagement, results indicate that the high location-high time entropy (i.e. multiple times, multiple locations) was the segment with lowest success when compared with other students. Statistical and modeling results also found that female students tended to learn at fixed or few locations resulting in the highest performance scores on the final exam. The primary implication is that female students tend to be successful because they study in fewer locations, and all students who study at consistent times outperform those with more varied time patterns. Existing brain research supports the findings on gender differences in learning performance and spatio-temporal characteristics.

Title: Design of pedestrian detectors using combinations of scale spaces and classifiers Abstract: With the increasing demand for surveillance applications, pedestrian detection has been a topic of interest for many researchers in recent time. The quality of a pedestrian detector is decided in terms of detection accuracy and rate of detection. This paper presents new pedestrian detectors based on two types of classifiers, linear support vector machine and cascade of boosted classifier. These classifiers are trained by using a feature set comprising of the histogram of oriented gradients and dense local difference binary features. Both the image pyramid and non-linear scale space are used to detect pedestrians of various sizes. In order to combine the benefits of the two classifiers, a new two-stage detection scheme is also presented. The detection accuracies of the proposed detectors are studied in terms of miss-rate versus false positive per image and miss-rate versus false positive per window. The performances of the detectors are also compared with the performances of existing detectors of similar type.

Title: Enhanced data security in MANET using trust-based Bayesian statistical model with RSSI by AOMDV Abstract: A Mobile Ad-hoc Network (MANET) is a form of mobile node which is known for its random movement and communication to other nodes without the need for any infrastructure. The Bayesian statistical model helps a routing protocol in MANET in getting a high mobility trusted node that relies on the outcome of the prediction. Fuzzy logic has been of immense use for trust in the calculation of RSSI value that helps finding a reliable and trusted route from source to destination. This innovative proposal that uses fuzzy logic in AOMDV is referred to as the Fuzzy trust with RSSI in Ad hoc On-demand Multipath Distance Vector (FT-AOMDV). It is known for its help in finding a reliable path in mobile Ad Hoc Networks. Simulation was conducted using a NS-2 network simulator under various scenarios that include varying node speeds and varying malicious nodes for the evaluation of the efficiency of the protocols. Metrics that include delivery ratio, end-to-end delay, routing packet overhead, and throughput have been used in the comparison of FT-AOMDV with AOMDV, TB-AOMDV1, and TB-AOMDV2.

Title: Efficient colour filter array demosaicking with prior error reduction Abstract: A single sensor camera captures only one intensity value in a pixel known as the raw/unprocessed image. The raw or incomplete image is reconstructed to a full colour image by the process called demosaicking. The proposed work introduces an error efficient demosaicking algorithm. The efficient prior error reduction technique helps to obtain better results. The demosaicked green image is a guide to the residual demosaicking process. The conventional demosaicking is replaced by the residual demosaicking. The standard Kodak and McMaster datasets are employed in the experimental analysis. The proposed methodology has produced optimal performance with reduced error compared to the conventional demosaicking algorithms.

Title: Deep neural network based two-stage Indian language identification system using glottal closure instants as anchor points Abstract: This paper presents a two-stage Indian language identification (TS-LID) system which is made up of a tonal/non-tonal pre-classification and individual language identification modules. It studies the effectiveness of Mean Hilbert envelope coefficients (MHEC) and Mel-frequency cepstral coefficients (MFCCs), and their combinations with prosody in TS-LID context. Both glottal closure instants (GCIs)-based approaches and the block processing (BP) approach have been explored. It also explores different types of analysis units, such as whole utterance and syllable. Various state-of-art modeling techniques have been analyzed in this work. Experiments have been carried out for the NIT Silchar language database (NITS-LD) and OGI-Multilingual database (OGI-MLTS). The results suggest that at the pre-classification stage, for NITS-LD, the deep neural network (DNN) with syllable-level features, using GCI-based approaches, provides the highest accuracies of 90.6%, 85% and 81.3% for 30 s, 10 s and 3 s test data respectively. The GCI-based approaches outperform the BP method by as much as 7.5%, 6.2%, and 5.7%. The pre-classification module helps to improve the performance of the LID system by as much as 5.7%, 4.4% and 2.2% for 30 s, 10 s and 3 s test data respectively. The corresponding improvements for OGI-MLTS database are 7.4%, 6.8%, and 5%.

Title: Ant colony optimization-induced route optimization for enhancing driving range of electric vehicles Abstract: Electric vehicles (EVs) are the emerging solution for pollution-free transportation systems in the modern era. Battery-operated motor-powered electric vehicles serve the purpose of transportation with in-time recharging ability. Routing and traversing locations using EVs demand optimal route selection for retaining delay and power of the vehicle. This manuscript proposes a fitness-ant colony optimization (FACO)-based route optimization for improving the driving range of EVs. FACO works in two phases: conditional route discovery and range-sustained traversing to control delay and to deprive EV failures because of earlier power drain. In range-sustained phase, the fitness of the traversing route is framed by considering the inputs of power and travel time of the EV for ensuring construction of optimal touring paths. The EVs are directed to traverse the paths defined through ACO, after which the available paths are further attuned to identify the most efficient route depending on the braking and battery power of the vehicle. This optimization balances both power and its variants along with travel time for improving the driving range of the EV before it actually drains out. The optimization technique is assessed using arbitrary road and delivery point simulation with real-time configurations to demonstrate the effectiveness of the proposed method. Experimental results demonstrate the consistency of the proposed FACO by increasing driving distance and delivery point visits. This optimization also achieves lesser power depletion retaining higher charging level with lesser waiting time.

Title: Online students' LMS activities and their effect on engagement, information literacy and academic performance Abstract: The purpose of this study was to examine online students' LMS activities and the effect on their engagement, information literacy, and academic performance. The participants of the study were 65 undergraduate students enrolled to an online "Computer Literacy" course. Cluster analysis was performed on the log data gathered from LMS activities, and participation levels were grouped according to two levels, as high participation and low participation. Multivariate analysis of variance (MANOVA) revealed that LMS participation levels could play an important role on student academic performance and engagement, but not for student information literacy. Closely monitoring student participation levels can help instructors determine students' needs and support learning accordingly. It can be stated that high levels of student participation enhance students' engagements to online courses. Thus, learning difficulties in online learning environments can be prevented. These findings may have implications for students' online learning processes, and also for instructional designs as they play an important role in enhancing students' success in online learning environments.

Title: Energy-efficient and delay-aware multitask offloading for mobile edge computing networks Abstract: Mobile edge computing (MEC) is a recent technology that intends to free mobile devices from computationally intensive workloads by offloading them to a nearby resource-rich edge architecture. It helps to reduce network traffic bottlenecks and offers new opportunities regarding data and processing privacy. Moreover, MEC-based applications can achieve lower latency level compared to cloud-based ones. However, in a multitask multidevice context, the decision of the part to offload becomes critical. Actually, it must consider the available communication resources, the resulting delays that have to be met during the offloading process, and particularly, both local and remote energy consumption. In this paper, we consider a multitask multidevice scenario where smart mobile devices retain a list of heavy offloadable tasks that are delay constrained. Therefore, we formulated the corresponding optimization problem, and we derive an equivalent multiple-choice knapsack problem formulation. Because of the short decision time constraint and the NP-hardness of the obtained problem, the optimal solution implementation is infeasible. Hence, we propose a solution that provides, in pseudopolynomial time, the optimal or near-optimal solutions depending on the problem's settings. In order to evaluate our solution, we carried out a set of simulation experiments to evaluate and compare the performances of the different components of this solution. Finally, the obtained results in terms of execution's time as well as energy consumption are satisfactory and very encouraging.

Title: Deep learning–based prediction model of occurrences of major adverse cardiac events during 1-year follow-up after hospital discharge in patients with AMI using knowledge mining Abstract: Traditional regression-based approaches do not provide good results in diagnosis and prediction of occurrences of cardiovascular diseases (CVD). Therefore, the goal of this paper is to propose a deep learning–based prediction model of occurrence of major adverse cardiac events (MACE) during the 1, 6, 12 month follow-up after hospital admission in acute myocardial infarction (AMI) patients using knowledge mining. We used the Korea Acute Myocardial Infarction Registry (KAMIR) dataset, a cardiovascular disease database registered in 52 hospitals in Korea between 1 January, 2005, and 31 December, 2008. Among 14,885 AMI patients, 10,813 subjects in age from 20 to 100 years with the 1-year follow-up traceability without coding errors were finally selected. For our experiment, the training/validation/test dataset split is 60/20/20 by random sampling without replacement. The preliminary deep learning model was first built by applying training and validation datasets and then a new preliminary deep learning model was generated using the best hyperparameters obtained from random hyperparameter grid search. Lastly, the preliminary prediction model of MACE occurrences in AMI patients is evaluated by test dataset. Compared with conventional regression-based models, the performances of machine/deep learning–based prediction models of the MACE occurrence in patients with AMI, including deep neural network (DNN), gradient boosting machine (GBM), and generalized linear model (GLM), are also evaluated through a matrix with sensitivity, specificity, overall accuracy, and the area under the ROC curve (AUC). The prediction results of the MACE occurrence during the 1, 6, and 12-month follow-up in AMI patients were the AUC of DNN (1 M 0.97, 6 M 0.94, 12 M 0.96), GBM (0.96, 0.95, 0.96), and GLM (0.76, 0.67, 0.72) in machine learning–based models as well as GRACE (0.75, 0.72, 0.76) in regression model. Compared with previous models, our deep learning–based prediction models significantly had the accuracy of 95% or higher and outperformed all machine learning and regression-based prediction models. This paper was the first trial of deep learning–based prediction model of the MACE occurrence in AMI clinical data. We found that the proposed prediction model applied different risk factors except the attribute “age” by using knowledge mining and directly used the raw data as input.

Title: Stability of feature selection algorithm: A review Abstract: Feature selection technique is a knowledge discovery tool which provides an understanding of the problem through the analysis of the most relevant features. Feature selection aims at building better classifier by listing significant features which also helps in reducing computational overload. Due to existing high throughput technologies and their recent advancements are resulting in high dimensional data due to which feature selection is being treated as handy and mandatory in such datasets. This actually questions the interpretability and stability of traditional feature selection algorithms. The high correlation in features frequently produces multiple equally optimal signatures, which makes traditional feature selection method unstable and thus leading to instability which reduces the confidence of selected features. Stability is the robustness of the feature preferences it produces to perturbation of training samples. Stability indicates the reproducibility power of the feature selection method. High stability of the feature selection algorithm is equally important as the high classification accuracy when evaluating feature selection performance. In this paper, we provide an overview of feature selection techniques and instability of the feature selection algorithm. We also present some of the solutions which can handle the different source of instability.

Title: Evolution of recommender paradigm optimization over time Abstract: In the past few decades recommender system has reshaped the way of information filtering between websites and the users. It helps in identifying user interest and generates product suggestions for the active users. This paper presents an enlightening analysis of various recommender system such as content-based, collaborative-based and hybrid recommendation techniques along with few optimization models that has been applied to improvise the parameters being considered by the aforementioned techniques. We explored 125 articles published from 1992 to 2019 in order to discuss the problems associated with the existing models. Various advantages and disadvantages of each recommendation model including the input methods has been elaborated. Critical review on research problems based on the explored techniques and future directions has also been covered.

Title: Detection and trace back of low and high volume of distributed denial-of-service attack based on statistical measures Abstract: DDoS attacks are rapidly growing bigger and upsetting online businesses than ever before. During 2012, DDoS attacks were performed on six US banks, viz, Bank of America, JP Morgan Chase, US Bancorp, Citigroup, and PNC Bank. The attacker(s) bombarded their targets with crest traffic of more than 60 gigabits per second. In 2013, Spamhaus, an antispam organization, was the victim for a DDoS attack of 300 gigabits per second. CyberBunker, a Dutch company, outwardly performed this attack on Spamhaus and got blacklisted after it. During 2014, the security company named Cloudflare was struck by 400 gigabits per second of traffic. Initially, the attack was aimed at one of Cloudflare's customers, but it later spread to the Cloudflare's own network. In 2018, a record-breaking attack was performed on GitHub, a popular code hosting website, with an impulsive blitz of traffic that marked the scales at 1.35 terabits per second. At this moment, from freelance websites to multinational banks, it sounds like nobody has enough online security. This paper presents a novel low and high-rate DDoS attack detection using statistical metrics. The NS3 simulation result shows that the proposed method detects the low and high volume of attacks in minimum time with reduced false positive and false negative than the existing method.

Title: Using robot-based practices to develop an activity that incorporated the 6E model to improve elementary school students' learning performances Abstract: This study used robot-based practices to develop an activity that incorporated the 6E model. The sixth-grade students learned interdisciplinary knowledge about how to use Arduino electronic components, microcontrollers, and hands-on tools to make a "Crab Robot." In addition, the students learned how to use Scratch programming language to control the robot and complete the "Crab Robot Crossing the Road" task. The study implemented a quasi-experimental design to examine whether the students who learned the robot-based activity using the 6E model acquired better learning motivation, learning performance, computational thinking ability, and hands-on ability than those who learned the activity through lectures. This study adopted purposive sampling to select 70 sixth-grade students from four classes, which were divided into the experimental group (6E model) and the control group (lectures). The experiment was conducted over a period of 18 weeks (for a total duration of 1,440 minutes). The results from the pretests-posttests showed that both groups of students improved their learning motivation, learning performance, computational thinking ability, and hands-on ability; however, the experimental group's scores were significantly better. More importantly, this study provides a pedagogy manuscript for instructors who want to teach mechatronics programs and programming design education.

Title: Multi-tier authentication schemes for fog computing: Architecture, security perspective, and challenges Abstract: Fog computing has revolutionized the computing domain by enabling resource sharing, such as online storage, and providing applications and software as services in near vicinity to the edge nodes through the Internet. Small- to large-sized companies, like Amazon, Google, Facebook, Twitter, and LinkedIn, have started switching to fog-computing-enabled infrastructures. Fog computing being distributed in nature and in near vicinity gives rises to security and privacy issues. Although mostly now a days, user identification is adopted via single sign-in process, such as simple password-based authentications, which is not a secure process. Several multi-tier authentication techniques are proposed to overcome single sign-in process limitations. In this article, we go through state-of-the-art schemes proposed over the period of 2011-2018 for multi-tier authentication, their weaknesses and security issues, and finally their solutions for fog-computing environment. We performed the comparison of available multi-tier authentication techniques based on three factors, ie, level of security, cost of deployment, and usability. Multi-tier authentication techniques are classified into categories in accordance with the aspects that are concerned with the authentication process. We are optimistic that this work will provide useful information to the researchers about the architectures of fog enabled systems and the underlying authentication models in a consolidated form.

Title: An overview of security and privacy in smart cities' IoT communications Abstract: Smart cities have brought significant improvements in quality of life and services to citizens and urban environments. They are fully enabled to control the physical objects in real time and provide intelligent information to citizens in terms of transport, healthcare, smart buildings, public safety, smart parking, and traffic system and smart agriculture, and so on. The applications of smart cities are able to collect sensitive information. However, various security and privacy issues may arise at different levels of the architecture. Therefore, it is important to be aware of these security and privacy issues while designing and implementing the applications. This paper highlights main applications of smart cities and addresses the major privacy and security issues in the architecture of the smart cities' applications. It also reviews some of the current solutions regarding the security and privacy of information-centric smart cities' applications and presents future research challenges that still need to be considered for performance improvement.

Title: Weighted Salp Swarm Algorithm and its applications towards optimal sensor deployment Abstract: Recent trends indicate the rapid growth of nature-inspired techniques in the field of optimization. Salp Swarm Algorithm (SSA) is a recently introduced stochastic algorithm that is inspired by the navigational capability and foraging behavior of Salps. However, classical SSA gives unsatisfactory results on higher dimension problems depicting poor convergence rate. The search process of SSA lacks exploration and exploitation resulting in convergence inefficiency. This paper proposes a strategy based on the weighted distance position update called Weighted Salp Swarm Algorithm (WSSA) to enhance the performance and convergence rate of the SSA. The proposed WSSA is validated using different benchmark functions and analyzed against seven different stochastic algorithms. The validation results confirmed enhanced performance and convergence rate of WSSA. Moreover, the proposed variant is applied for optimal sensor deployment task. WSSA approach is applied on probabilistic sensor model to maximize coverage and radio energy model to minimize energy consumption. This strategy is a trade-off between coverage and energy efficiency of the sensor network. It was observed that WSSA algorithm outperformed all the other stochastic algorithms in optimizing coverage and energy efficiency of Wireless Sensor Network (WSN).

Title: Optimal floor planning in VLSI using improved adaptive particle swarm optimization Abstract: Floor planning is necessary to design the VLSI circuit. The complete computational characteristics of the manufactured chip are evaluated by floor planning process. It is the multi-objective problem in which different objectives are fulfilled at a time. Here, a new Interactive Self-Improvement based Adaptive Particle Swarm Optimization (ISI-APSO) technique is proposed to enhance the exploration efficiency and accuracy than convolutional PSO. Within less computation time the proposed ISI-APSO technique attains best global search throughout the space. The simulation results show that the proposed ISI-APSO algorithm achieves better performance than other heuristic algorithms in exploring efficiency and speed of convergence. In order to place the whole modules and their internally connected wire lengths, the Multi-objective optimization method is utilized. Therefore the necessary layout area is minimized. Moreover, the implemented results demonstrate the importance of the proposed algorithm with respect to the robust performance.

Title: A modified variable neighborhood search algorithm for manufacturer selection and order acceptance in distributed virtual manufacturing network Abstract: In this paper, we investigate a manufacturer selection and composition problem for Distributed Virtual Manufacturing Network (DVMN) with order acceptance and scheduling of deteriorating jobs, where potential manufacturers include proprietary plants and outsourced co-manufacturers. In such a problem, at the beginning of planning horizon, a manufacturing company (MC) receives a series of order requests from its customers. Due to the limited production capacity, the MC is not able to fulfill all the orders requested by the customers on time. To satisfy the order demand as much as possible, one of the MC's choices is to establish a collaborative manufacturing platform (i.e., DVMN) that leverages the advantages of a social manufacturing network to expand their production capabilities. However, the increase in production capacity does not mean that all orders from the customers are able to be accepted and produced. In fact, standing at the profit-maximizing point of view, the order requests that do not bring profits to the company need to be rejected at the beginning of planning horizon. For the accepted orders, the MC will produce a part of them by its plants in the DVMN, and the remaining orders will be outsourced to other co-manufacturers in the DVMN. Based on some structural properties, we develop a novel Reduced Variable Neighborhood Search (RVNS) algorithm incorporated multiple random mutations neighborhood structures (NSs-MRM) to determine the composition of the DVMN, and further determine which orders should be rejected, which accepted orders should be processed by the company-owned plants in the DVMN, and which accepted orders should be outsourced to the co-manufacturers in the DVMN. In particular, the processing sequence of the orders produced in the company-owned plants are also determined. Finally, a reasonable composition of the DVMN and corresponding order acceptance as well as order scheduling on each member of the DVMN are given. To evaluate the proposed algorithm, a series of computational experiments are conducted, and the experimental results show that the proposed algorithm outperforms other existing baseline algorithms in solving both small-scale and large-scale problem instances.

Title: Examining the impact of head-mounted display virtual reality on the science self-efficacy of high schoolers Abstract: Traditional science, technology, engineering, and mathematics (STEM) education is sometimes criticized for lacking approaches to present real-world practices and phenomena beyond naked eyes. Head-mounted display virtual reality (HMD VR) provides opportunities to solve this issue. However, little is known about the impact of this approach on student's self-efficacy in science. This study is to address this knowledge gap. Sixty-six 11th grade students were recruited to participate in an HMD VR learning activity. Half of these students filled in a science self-efficacy questionnaire before the VR learning activity, and the others filled in it after the activity. The study compared (1) students' science self-efficacy between these two conditions and (2) students' post-activity science self-efficacy among different science-class grading score groups. Results showed that the change of students' science self-efficacy was not significant after the learning activity and the differences among most science-class grading score groups were small. After the results were analyzed, the capability of affording gestures and physical movement was recognized as an important factor that determined whether an HMD VR learning environment could significantly enhance students' science self-efficacy; educators were suggested to not use science class scores to predict students' potential and future achievements in science.

Title: Solving the index tracking problem: a continuous optimization approach Abstract: Investing vast amounts of money with the goal of fostering medium to long-term growth in returns is a challenging task in financial optimization. A method might be mirroring the market index as closely as possible by choosing from the stocks that make up the index. This approach is known as index tracking and the objective of this paper is to address this problem in order to solve it by means of mathematical programming techniques. In particular, we are interested in investigating the index tracking problem (ITP) as a mixed integer linear program in presence of some real-world constraints known as cardinality constraints as well as transaction costs. These ITP models are NP-hard, and consequently, difficult to solve by classical exact methods even for medium-sized instances. In order to overcome this issue, we propose a method based on nonconvex programming techniques. More precisely, we reformulate the problem as a difference of convex functions (DC) program and solve it by means of an approach known as DC algorithm. In order to evaluate the performance of the proposed algorithm, we conducted numerical experiments using benchmark instances. The results of the algorithm are compared with those provided by the state-of-the-art MILP solver Gurobi. The numerical results confirm the efficiency of the method in solving the ITP.

Title: Android malware detection through generative adversarial networks Abstract: Mobile and cell devices have empowered end users to tweak their cell phones more than ever and introduce applications just as we used to with personal computers. Android likewise portrays an uprise in mobile devices and personal digital assistants. It is an open-source versatile platform fueling incalculable hardware units, tablets, televisions, auto amusement frameworks, digital boxes, and so forth. In a generally shorter life cycle, Android also has additionally experienced a mammoth development in application malware. In this context, a toweringly large measure of strategies has been proposed in theory for the examination and detection of these harmful applications for the Android platform. These strategies attempt to both statically reverse engineer the application and elicit meaningful information as features manually or dynamically endeavor to quantify the runtime behavior of the application to identify malevolence. The overgrowing nature of Android malware has enormously debilitated the support of protective measures, which leaves the platforms such as Android feeble for novel and mysterious malware. Machine learning is being utilized for malware diagnosis in mobile phones as a common practice and in Android distinctively. It is important to specify here that these systems, however, utilize and adapt the learning-based techniques, yet the overhead of hand-created features limits ease of use of such methods in reality by an end user. As a solution to this issue, we mean to make utilization of deep learning-based algorithms as the fundamental arrangement for malware examination on Android. Deep learning turns up as another way of research that has bid the scientific community in the fields of vision, speech, and natural language processing. Of late, models set up on deep convolution networks outmatched techniques utilizing handmade descriptive features at various undertakings. Likewise, our proposed technique to cater malware detection is by design a deep learning model making use of generative adversarial networks, which is responsible to detect the Android malware via famous two-player game theory for a rock-paper-scissor problem. We have used three state-of-the-art datasets and augmented a large-scale dataset of opcodes extracted from the Android Package Kit bytecode and used in our experiments. Our technique achieves F1 score of 99% with a receiver operating characteristic of 99% on the bytecode dataset. This proves the usefulness of our technique and that it can generally be adopted in real life.

Title: Glowworm swarm based fuzzy classifier with dual features for speech emotion recognition Abstract: Nowadays, a great attention is focusing on the study of the emotional content in speech signals since the speech signal is one of the quickest and natural tactic to communicate among humans, and thus, many systems have been suggested to recognize the emotional content of a spoken utterance. This paper schemes two contributions: Gender recognition and Emotion recognition. In the first contribution, there has two phases: feature extraction and classification. The pitch feature is extracted and given for classification using k-Nearest Neighboring classifier. In the second contribution, features like Non-Negative Matrix Factorization and pitch are extracted and given as the input to Adaptive Fuzzy classifier to recognize the respective emotions. In addition, the limits of membership functions are optimally chosen using a renowned optimization algorithm namely glowworm swarm optimization (GSO). Thus the proposed adaptive Fuzzy classifier using GSO is termed as GSO-FC. The performance of proposed model is compared to other conventional algorithms like Grey Wolf Optimization, FireFly, Particle Swarm Optimization, Artificial Bee Colony and Genetic Algorithm in correspondence with varied performance measures like Accuracy, Sensitivity, Specificity, Precision, False positive rate, False negative rate, Negative Predictive Value, False Discovery Rate, F1 Score and Mathews correlation coefficient.

Title: Generalized PVO-based dynamic block reversible data hiding for secure transmission using firefly algorithm Abstract: In this paper, we proposed a novel generalized pixel value ordering-based reversible data hiding using firefly algorithm (GPVOFA). The sequence of minimum and maximum number pixels value has been used to embed the secret data while prediction and modification are held on minimum, and the maximum number of pixel blocks is used to embed the secret data into multiple bits. The host image is divided into the size of noncoinciding dynamic blocks on the basis of firefly quadtree partition, whereas rough blocks are divided into a larger size; moreover, providing more embedding capacity used small flat blocks size and optimal location in the block to write the information. Our proposed method becomes able to embed large data into a host image with low distortion. The rich experimental results are better, as compared with related preceding arts.

Title: Frequency based tactile rendering method for pin-array tactile devices Abstract: In interactive Internet of Things (IoT) environment, a solenoid-based tiny pin-array tactile device can become a key module to haptically simulate a surface of interconnected objects. The solenoid-based pin-array tactile device creates a large enough force and stroke to stimulate human skin and generates a wide frequency range. The solenoid-based tiny pin-array tactile device, however, brings a new issue of controlling the pin's stroke of the tactile device. To overcome the limitation of the solenoid-based device, a new tactile rendering method is needed. In the proposed tactile rendering method, we control the operating frequency of pins, instead of controlling their stroke, to haptically simulate the surface of interconnected objects. Our experiments demonstrate that our proposed method with a pin-array tactile device is suitable for simulating the surface of interconnected objects.

Title: Ensemble of fast learning stochastic gradient boosting Abstract: Boosting is one of the most popular and powerful learning algorithms. However, due to its sequential nature in model fitting, the computational time of boosting algorithm can be prohibitive for big data analysis. In this paper, we proposed a parallel framework for boosting algorithm, called Ensemble of Fast Learning Stochastic Gradient Boosting (EFLSGB). The proposed EFLSGB is well suited for parallel execution, and therefore, can substantially reduce the computational time. Analysis of simulated and real datasets demonstrates that EFLSGB achieves highly competitive prediction accuracy in comparison with gradient tree boosting.

Title: Higher-order asymptotic refinements in the multivariate Dirichlet regression model Abstract: The likelihood ratio test statistic provides the basis for testing inference on the regression parameters in the class of multivariate Dirichlet regression models, which is very useful in modeling multivariate positive observations summing to one. We focus on the small-sample case, where the reference chi-squared distribution gives a poor approximation to the true null distribution of the likelihood ratio statistic. Our simulation results suggest that the likelihood ratio test tends to be extremely liberal when the sample size is small. We derive a general Bartlett correction factor in matrix notation for the likelihood ratio test statistic, which reduces the size distortion of the test, and also consider a bootstrap-based Bartlett correction. We also employ the Skovgaard's adjustment to the likelihood ratio statistic. We numerically compare the proposed tests with the usual likelihood ratio test. Our simulation results suggest that the proposed corrected tests can be interesting alternatives to usual likelihood ratio test since they lead to very accurate inference even for very small samples. We also present an empirical application for illustrative purposes.

Title: Degree Sum Condition for the Existence of Spanning k-Trees in Star-Free Graphs Abstract: For an integer k >= 2, a k-tree T is defined as a tree with maximum degree at most k. If a k-tree T spans a graph G, then T is called a spanning k-tree of G. Since a spanning 2-tree is a Hamiltonian path, a spanning k-tree is an extended concept of a Hamiltonian path. The first result, implying the existence of k-trees in star-free graphs, was by Caro, Krasikov, and Roditty in 1985, and independently, Jackson and Wormald in 1990, who proved that for any integer k with k >= 3, every connected K-1(,k)-free graph contains a spanning k-tree. In this paper, we focus on a sharp condition that guarantees the existence of a spanning k-tree in K-1(,k)+1-free graphs. In particular, we show that every connected K-1(,k)+1-free graph G has a spanning k-tree if the degree sum of any 3k - 3 independent vertices in G is at least |G| - 2, where |G| is the order of G.

Title: Hotelling T-2 control chart based on bivariate ranked set schemes Abstract: Univariate control charts under ranked set schemes are used to continuously monitor two or more related characteristic independently. In this article we have offered Hotelling T-2 control chart based on different bivariate ranked set schemes for monitoring two related characteristic simultaneously instead of independently. In this regard, we have proposed control chart statistic and control limits under bivariate ranked set schemes for handling two characteristics. Furthermore, we have computed empirical run length distribution and its properties to see the effect or role of different factors. We have found that proposed Hotelling T-2 control chart under ranked set schemes perform outstanding as compared to classical Hotelling T-2 control chart under simple random sampling in general. Moreover, proposed Hotelling T-2 control chart under double ranked set schemes is declared more efficient relative to single ranked set schemes. Among single and double ranked set schemes, proposed control chart under median ranked set schemes have uniformly high detection ability. Another feature of proposed control chart is to become special cases of the existing control charts when specific value of correlation between study variables and concomitant variable as well as sample size are considered. A real life example has included in which irrigation water monitored through the proposed control chart.

Title: Open Locating-Dominating Sets in Circulant Graphs Abstract: Location detection problems have been studied for a variety of applications including finding faults in multiprocessors, contaminants in public utilities, intruders in buildings and facilities, and for environmental monitoring using wireless sensor networks. In each of these applications, the system or structure can be modeled as a graph, and sensors placed strategically at a subset of vertices can locate and detect anomalies in the system. An open locating-dominating set (OLD-set) is a subset of vertices in a graph in which every vertex in the graph has a non-empty and unique set of neighbors in the subset. Sensors placed at OLD-set vertices can uniquely detect and locate disturbances in a system. These sensors can be expensive and, as a result, minimizing the size of the OLD-set is critical. Circulant graphs, a group of regular cyclic graphs, are often used to model parallel networks. We prove the optimal OLD-set size for a particular circulant graph using Hall's Theorem. We also consider the mixed-weight OLD-set introduced in [R.M. Givens, R.K. Kincaid, W. Mao and G. Yu, Mixed-weight open locating-dominating sets, in: 2017 Annual Conference on Information Science and Systems, (IEEE, Baltimor, 2017) 1-6] which models a system with sensors of varying strengths. To model these systems, we place weights on the vertices in the graph, representing the strength of a sensor placed at the corresponding location in the system. We study particular mixed-weight OLD-sets in cycles, which behave similarly to OLD-sets in circulant graphs, and show the optimal mixed-weight OLD-set size using the discharging method.

Title: Investigating the effect of real-time multi-peer feedback with the use of a web-based polling software on e-learners' learning performance Abstract: This study seeks to investigate the impact of a real-time multi-peer feedback process with the use of a web-based polling software on adult e-learners' learning performance. Two groups of participants with 30 members in each were involved in the experiment. In the two-and-a-half-hour experiment, the experimental group adopted a web-based polling software - Rain Classroom to carry out the real-time multi-peer feedback activities for the learning of some English infinitive verbs, while the control group was applied the real-time single-peer feedback mode in learning the same content. A pretest, a posttest, a delayed posttest, a questionnaire survey and a follow-up interview were employed as the instruments for the experiment. The results revealed that the multi-peer group significantly outperformed the single-peer group on the delayed posttest, though no significant difference was found in the posttest between the two groups. The questionnaire survey and the follow-up interview showed that the multi-peer method gained great popularity among the participants and could enhance learners' engagement in learning. Other related issues are also discussed.

Title: High fidelity based reversible data hiding using modified LSB matching and pixel difference Abstract: Owing to the inefficiency to hide large volume of secret data for the reversible data hiding (RDH) image steganography approaches, we propose two improved RDH based approaches, such as (1) improved dual image based least significant bit (LSB) matching with reversibility, and (2) n-rightmost bit replacement (n-RBR) and modified pixel value differencing (MPVD). The first approach extends the ability of LSB matching with reversibility using dual images. Whereas the second approach utilizes four identical cover images for secret data embedding using two phases, such as (1) n-rightmost bit replacement (n-RBR) and (2) modified pixel value differencing (MPVD). In the n-RBR phase, n bits of secret data are embedded in the pair of two neighboring pixels of the first two identical images, where 1 ≤ n ≤ 4. Correspondingly, the MPVD phase uses the third and fourth identical images for hiding the secret data. Experimental results with respect to peak signal-to-noise ratio (PSNR), embedding capacity (EC), structural similarity index (SSIM), and the comparative analysis with recently proposed state-of-art approaches exhibit the superiority of the proposed approach. Besides reversibility, the proposed approach ensures high fidelity to salt and pepper (S&P) noise, RS analysis, and pixel difference histogram (PDH) analysis.

Title: A two-stage intervened decision system with multi-state decision units and dynamic system configuration Abstract: This paper develops the performability and cost–benefit models for a two-stage intervened decision system with majority voting rule and binary input and output. The decision process of the system contains two stages: an inspection stage (stage 1) and a result submission stage (stage 2). During the first stage, each decision unit in the system will have multiple states and a supervisor will come to visit each unit and check its state for at most twice. The supervisor will conduct the first visit to each unit for certain. However, the behavior of the second visit to each unit will be determined by its state during the first visit. In addition, each decision unit may be removed from the system given certain states during each visit. Therefore the structure of system may change during the decision process. The units which are not removed during the first stage can submit the result at any time during the second stage. However, the performance of each remaining unit will be determined by the ending state of the first stage. Moreover, in order to improve the efficiency of the decision process, a check point is added to the second stage. The performability and cost–benefit models for this dynamic system are developed by considering the distribution of states at the end of the first stage. A three-step method will be proposed for model optimization. Some numerical examples for the three-step method will be presented. The proposed intervened decision system in this paper can be applied in many contexts such as financial investment, paper submission review and proposal evaluation, credit evaluation and loan application and product release and recall.

Title: Robust variable selection based on the random quantile LASSO Abstract: In this paper, we study robust variable selection problem by combining the idea of quantile regression and random LASSO. A two-step algorithm is proposed to solve the proposed optimization problem. In the first step, we use bootstrap samples and variable subsets to estimate the importance of each variable. In the second step, the importance measures are used in generating variable subsets and then the adaptive quantile LASSO is applied to reduce the bias of estimators. The proposed method is robust and can handle the situation of highly-correlated variables. Meanwhile, the number of selected variables is no longer limited by the sample size. Simulation studies indicate that the proposed method has good robustness and better performance when the error term is heavy-tailed and there are highly correlated variables. Finally, we apply the proposed methodology to analyze a real data. The results reveal that the propose has better the predictive ability.

Title: Node selection and utility maximization for mobile edge computing-driven IoT Abstract: Optimal user association and resource utilization entails a challenging problem of high latency in mobile edge computing (MEC)-driven Internet of Things (IoT) applications. Increased power consumption is another aspect that requires attention, specifically for the systems that involve huge number of users (IoT nodes) and computing devices (cloudlets/MEC nodes/fog nodes). Though MEC/fog networks are designed for low latency and low transmit power connections, yet the substantial increase in IoT devices signify the need to redesign it for future demands. In this article, we formulate an optimization problem related to quality of service-driven node selection and utility maximization subject to power and workload constraints for applications that require very low latency. Outer approximation algorithm is a proven technique in the field of optimization theory and is scalable with number of nodes. A distributed self-converging algorithm based on the outer approximation algorithm is presented in this work, which proved to be efficient for the problem formulated. Extensive simulations are done to validate the numerical results. This work concludes by comparing results of outer approximation method with matching theory to validate its effectiveness.

Title: Multi-objective optimization for stochastic failure-prone job shop scheduling problem via hybrid of NSGA-II and simulation method Abstract: Production scheduling and reliability of machinery are prominent issues in flexible manufacturing systems that are led to decreasing of production costs and increasing of system efficiency. In this paper, multiobjective optimization of stochastic failure-prone job shop scheduling problem is sought wherein that job processing time seems to be controllable. It endeavours to determine the best sequence of jobs, optimal production rate, and optimum preventive maintenance period for simultaneous optimization of three criteria of sum of earliness and tardiness, system reliability, and energy consumption. First, a new mixed integer programming model is proposed to formulate the problem. Then, by combining of simulation and NSGA-II algorithm, a new algorithm is put forward for solving the problem. A set of Pareto optimal solutions is achieved through this algorithm. The stochastic failure-prone job shop with controllable processing times has not been investigated in the earlier research, and for the first time, a new hedging point policy is presented. The computational results reveal that the proposed metaheuristic algorithm converges into optimal or near-optimal solution. To end, results and managerial insights for the problem are presented.

Title: A SDN-based fine-grained measurement and modeling approach to vehicular communication network traffic Abstract: In smart cities, a large number of vehicles are connected into an intelligent transportation system and share information via the vehicular communication network (VCN). Accurate, fine-grained, and comprehensive traffic measurements are very crucial for the controller's decision making in software-defined networking (SDN) in VCN. Fine-grained traffic measurements can accurately portray network behaviors for VCN. However, this will increase a large amount of measurement overhead. Therefore, how to effectively obtain accurate and fine-grained traffic is a huge challenge for VCN. To the end, this paper proposes a novel accurate and SDN-based fine-grained traffic measurement approach to obtain comprehensive traffic in VCN. Firstly, based on SDN architecture, we exploit the pull-based sampling mechanism to quickly obtain coarse-grained traffic measurement values. Secondly, based on the matrix completion theory, we use both interpolation and optimization methods to obtain fine-grained traffic measurements. Thirdly, the optimization model and detailed algorithm are proposed to attain accurate traffic. Finally, we conduct a larger number of simulations to validate the measurement approach proposed in this paper. Simulation results show that our approach exhibits better performance and is promising.

Title: A test statistic with a ranking method based on the Jeffreys divergence measure Abstract: Nonparametric two-sample tests are important statistical procedures in many scientific fields. The test statistic has been derived from the Kullback-Liebler divergence between two empirical distribution functions. This study modifies a nonparametric two-sample test statistic using a ranking method. The modified test statistic is shown to be based on the Jeffreys divergence measure. The exact critical value of the proposed test statistic is derived for small sample sizes. Simulations are used to investigate the power of the proposed test statistic for the location alternative and for any difference between distributions, with various population distributions, for small sample sizes.

Title: Subgroup discovery in MOOCs: a big data application for describing different types of learners Abstract: The aim of this paper is to categorize and describe different types of learners in massive open online courses (MOOCs) by means of a subgroup discovery (SD) approach based on MapReduce. The proposed SD approach, which is an extension of the well-known FP-Growth algorithm, considers emerging parallel methodologies like MapReduce to be able to cope with extremely large datasets. As an additional feature, the proposal includes a threshold value to denote the number of courses that each discovered rule should satisfy. A post-processing step is also included so redundant subgroups can be removed. The experimental stage is carried out by considering de-identified data from the first year of 16 MITx and HarvardX courses on the edX platform. Experimental results demonstrate that the proposed MapReduce approach outperforms traditional sequential SD approaches, achieving a runtime that is almost constant for different courses. Additionally, thanks to the final post-processing step, only interesting and not-redundant rules are discovered, hence reducing the number of subgroups in one or two orders of magnitude. Finally, the discovered subgroups are easily used by courses' instructors not only for descriptive purposes but also for additional tasks such as recommendation or personalization.

Title: n-Grams exclusion and inclusion filter for intrusion detection in Internet of Energy big data systems Abstract: The advent of Internet of Energy (IoE) and the seamless integration of grid operators, power generators, distributors, sensors, and end users promise more efficient use of energy. However, the IoE will inherit the vulnerabilities from all of the integrated systems, and this raises concerns for trust and privacy. The evolving complexity and increased speed of network-based attacks emphasizes the need for an efficient intrusion detection system. Consequently, with the emergence of new attacks and the increasing number of signatures, traditional signature-based intrusion detection systems cannot both sift through big data and meet high network speeds. Detection performance severely deteriorates when matching hundreds of gigabits per second to the growing number of attack signatures. Given that pattern matching takes up to 60% of the overall intrusion detection time, this paper presents a new and fast software-based pattern matching system, Exscind. It proposes an exclusion-inclusion filter to preclude clean traffic before having to do expensive pattern matching. Additionally, if the traffic is malicious, the system only matches against a subset of signatures that have a high probability of being a match. We extensively evaluate the system's performance and conclude that using 6-grams signature prefix provides the best speedup and memory consumption with negligible false positives and linear scaling. We report a best-case speedup of 6.5 times for normal traffic and 1.53 times for the worst possible scenario. For best-case normal traffic, Exscind skips pattern matching for 98.36% of the packets.

Title: SSPA: an effective semi-supervised peer assessment method for large scale MOOCs Abstract: Peer assessment has become a primary solution to the challenge of evaluating a large number of students in Massive Open Online Courses (MOOCs). In peer assessment, all students need to evaluate a subset of other students' assignments, and then these peer grades are aggregated to predict a final score for each student. Unfortunately, due to the lack of grading experience or the heterogeneous grading abilities, students may introduce unintentional deviations in the evaluation. This paper proposes and implements a semi-supervised peer assessment method (SSPA) that incorporates a small number of teacher's gradings as ground truth, and uses them to externally calibrate the procedure of aggregating peer grades. Specifically, each student's grading ability is directly (if students have common peer assessments with teacher) or indirectly (if students have no common peer assessments with teacher) measured with the grading similarity between the student and teacher. Then, SSPA utilizes the weighted aggregation of peer grades to infer the final score of each student. Based on both real dataset and synthetic datasets, the experimental results illustrate that SSPA performs better than the existing methods.

Title: A multivariant secure framework for smart mobile health application Abstract: Wireless sensor network enables remote connectivity of technological devices such as smart mobile with the internet. Due to its low cost as well as easy availability of data sharing and accessing devices, the Internet of Things (IoT) has grown exponentially during the past few years. The availability of these devices plays a remarkable role in the new era of mHealth. In mHealth, the sensors generate enormous amounts of data and the context-aware computing has proven to collect and manage the data. The context aware computing is a new domain to be aware of context of involved devices. The context-aware computing is playing a very significant part in the development of smart mobile health applications to monitor the health of patients more efficiently. Security is one of the key challenges in IoT-based mHealth application development. The wireless nature of IoT devices motivates attackers to attack on application; these vulnerable attacks can be denial of service attack, sinkhole attack, and select forwarding attack. These attacks lead intruders to disrupt the application's functionality, data packet drops to malicious end and changes the route of data and forwards the data packet to other location. There is a need to timely detect and prevent these threats in mobile health applications. Existing work includes many security frameworks to secure the mobile health applications but all have some drawbacks. This paper presents existing frameworks, the impact of threats on applications, on information, and different security levels. From this line of research, we propose a security framework with two algorithms, ie, (i) patient priority autonomous call and (ii) location distance based switch, for mobile health applications and make a comparative analysis of the proposed framework with the existing ones.

Title: Imperialist competitive algorithm-based deep belief network for food recognition and calorie estimation Abstract: The vulnerabilities of the health issues have resulted in the alternative to manage the situation, which ensures the betterment in life. The dietary assessment stands as an effective solution for most of the health vulnerabilities and the automatic assessment takes-off the manual procedure of assessing the food intake. This paper introduces an automatic method of dietary assessment by proposing the Imperialist Competitive Algorithm (IpCA)-based Deep Belief Network (IpCA-DBN) for food category recognition and the calorie estimation of the food. Initially, the food image is pre-processed and subjected to the segmentation process, which is done by the Bayesian Fuzzy Clustering. Then, the features, such as shape, color histogram, wavelet, scattering transform features are generated from the optimal segments. Finally, these features are fed to the IpCA-DBN for recognizing the food category and estimating the calorie of the food. The experimentation performed using the UNIMIB2016 dataset enables the effective analysis of the proposed method in terms of the metrics, such as Macro Average Accuracy (MAA), Standard Accuracy (SA), and Mean Square Error (MSE). The analysis proves that the proposed method outperforms the existing methods and attained 0.9643 for MAA, 0.9877 for SA, and 1816.9 for MSE.

Title: Energy-efficient source location privacy protection for network lifetime maximization against local eavesdropper in wireless sensor network (EeSP) Abstract: Wireless sensor networks can be deployed in harsh environments for monitoring purposes. In such environments where there is a lack of human presence, the network may face severe security and privacy issues due to unauthorized access of a powerful attacker. The attacker may exploit the contextual information to locate the position of the asset. We propose a scheme for protecting the location of the source node, which consumes less energy hence maximizing network lifetime. The proposed scheme has the inspiration of black and white regions of the chessboard, which represent sleep and active regions in the wireless sensor network. This alternation plays its role in reducing the total energy consumption along misguiding the eavesdropper from finding the source node location. Our scheme uses dynamic cluster head selection to further add to location privacy. The results show that we have improved the safety period by a maximum of 131% and a minimum of 17% than that of the compared techniques. Similarly, a maximum of 60% reduction in energy consumption and a 20% reduction in latency is recorded by our proposed scheme. These results make our scheme suitable for application requiring high-safety period along maximizing network lifetime.

Title: Design of optimized energy system based on active energy-saving technologies in very low-energy smart buildings Abstract: The International Energy Agency proclaims that the energy utilization by commercial buildings makes up 28% of the world's total energy consumption. This research is aimed at finding various optimum ways to minimize energy consumption for existing commercial buildings. A three-layer framework is proposed to achieve active energy saving and to transform an existing building into very low-energy building. The building considered is the postgraduate studies research lab that can have 15 workstations. The building site (envelop) is made scalable by dividing the site area into zones and each zone is monitored via a sensor node, which monitors occupant behavior and acts as a controlling agent between the source and the load. The use of photovoltaic as renewable generation is a sustainable and environment-friendly way and thus, is added to the existing building to conserve energy. A 6 kW grid-tied monocrystalline photovoltaic-based renewable energy system has been added to the building to meet energy requirements and envisage its aftereffect on the energy conservation. The energy-efficient appliances have been used as loads to minimize the energy consumption. The installed system has a payback period of 6 years, while the proposed framework can achieve maximum energy efficiency of 90%.

Title: Programming trajectories analytics in block-based programming language learning Abstract: Block-based programing languages (BBPL) provide effective scaffolding for K-12 students to learn computational thinking. However, the output-based assessment in BBPL learning is insufficient as we can not understand how students learn and what mistakes they have had. This study aims to propose a data-driven method that provides insight into students' problem-solving process in a game-based BBPL practice. Based on a large-scale programing dataset generated by 131,770 students in solving a classical maze game with BBPL in Hour of Code, we first conducted statistical analysis to extract the most common mistakes and correction trajectories students had. Furthermore, we proposed a novel program representation method based on tree edit distance of abstract syntax tree to represent students' programing trajectories, then applied a hierarchical agglomerative clustering algorithm to find the hidden patterns behind these trajectories. The experimental results revealed four qualitatively different clusters: quitters, approachers, solvers and knowers. The further statistical analysis indicated the significant difference on the overall performance among different clusters. This work provides not only a new method to represent students' programing trajectories but also an efficient approach to interpret students' final performance from the perspective of programing process.

Title: Visualization of knowledge map for monitoring knowledge diagnoses Abstract: In order to evaluate the level of knowledge acquired by students, this study presents the contents of the students' knowledge using a proposed novel method to monitor a knowledge map and to diagnose them. To do so, we propose a method to improve the accuracy of diagnoses by comparing the level of the surrounding knowledge related to the knowledge, rather than the level of the knowledge itself, to correct the evaluation value. Moreover, in visualizing the relationship between knowledge objects based on knowledge evaluation data using an ontological structure, the Deep Sparse Neural Network model of deep learning is applied, and the map is regarded as one neural network and is proposed to express the quantitative value using the weight instead of the qualitative concept of ontology. The proposed knowledge map visualization can monitor the relationship and relevance (weight) of the related knowledge level at a glance, so that it is possible to intuitively grasp the result with the quantitative knowledge diagnosis and to improve the efficiency of knowledge evaluation and the accurate evaluation of knowledge diagnoses.

Title: Authorized Keyword Search over Outsourced Encrypted Data in Cloud Environment Abstract: For better data availability and accessibility while ensuring data secrecy, end-users often tend to outsource their data to the cloud servers in an encrypted form. However, this brings a major challenge to perform the search for some keywords over encrypted content without disclosing any information to unintended entities. This paper proposes a novel expressive authorized keyword search scheme rel...

Title: Pretesting strategies for homoscedasticity when comparing means. Their robustness facing non-normality Abstract: Traditional pretests to prove homoscedasticity (e.g., Levene's or Bartlett's test) before applying normal parametric techniques like the Student's t test should be avoided: Given the form of their null and alternative hypotheses, they are methodologically inadequate to this end and induce overall (regarding the full testing process, pretesting included) alterations in Type I Error Probability (TIEP). Under normality, not pretesting and always applying the Welch's test (instead of Student's test) or pretesting with the Wellek's equivalence test of heteroscedasticity irrelevance, are more affordable strategies. Here we investigate the robustness, in front of non-normality, of five strategies dealing with possible heteroscedasticity: Pretesting with Levene, F and Wellek's, and applying either the Student's or the Welch's test directly without pretest. Non-normality is simulated by varying degrees of skewness and kurtosis, for distributions inside the Fleishman's system. Robustness is measured in terms of overall TIEP proximity to the nominal significance level. The direct use of Welch's test or pretesting with the Wellek's test are still the most robust approaches, although high degrees of contamination of normality affect them considerably. All approaches improve with growing sample sizes and worsen with unbalancing, but differences still persist, in general favoring the above mentioned two approaches.

Title: Finding the polygon hull of a network without conditions on the starting vertex Abstract: Many real-life problems arising within the fields of wireless communication, image processing, combinatorial optimization, etc, can be modeled by means of Euclidean graphs. In the case of wireless sensor networks, the overall topology of the graph is not known because sensor nodes are often randomly deployed. One of the significant problems in this field is the search for boundary nodes. This problem is important in cases such as the surveillance of an area of interest, image contour reconstruction, graph matching problems, routing or clustering data, etc. In the literature, many algorithms are proposed to solve this problem, a recent one of which is the least polar-angle connected node (LPCN) algorithm and its distributed version D-LPCN, which are both based on the concept of a polar angle visit. An inconvenience of these algorithms is the determination of the starting vertex. In effect, the point with the minimum x-coordinate is a possible starting point, but it has to be known at the beginning, which considerably increases the algorithms' complexity. In this article, we propose a new method called RRLPCN (reset and restart with least polar-angle connected node), which is based on the LPCN algorithm to find the boundary vertices of a Euclidean graph. The main idea is to start the LPCN algorithm from an arbitrary vertex, and whenever it finds a vertex with an x-coordinate smaller than that of the starting one, LPCN is reset and restarted from this new vertex. The algorithm stops as soon as it visits the same edge for the second time in the same direction. In addition to finding the boundary vertices, RRLPCN also finds the vertex with minimum x-coordinate, which is the last starting point of our algorithm. The distributed version of the proposed algorithm, called D-RRLPCN, is then applied to boundary node detection in the wireless sensor network. It has been implemented using real sensor nodes (Arduino/XBee and TelosB). The simulation results have shown our algorithm to be very effective in comparison to other algorithms.

Title: Edge computing in smart health care systems: Review, challenges, and research directions Abstract: Today, patients are demanding a newer and more sophisticated health care system, one that is more personalized and matches the speed of modern life. For the latency and energy efficiency requirements to be met for a real-time collection and analysis of health data, an edge computing environment is the answer, combined with 5G speeds and modern computing techniques. Previous health care surveys have focused on new fog architecture and sensor types, which leaves untouched the aspect of optimal computing techniques, such as encryption, authentication, and classification that are used on the devices deployed in an edge computing architecture. This paper aims first to survey the current and emerging edge computing architectures and techniques for health care applications, as well as to identify requirements and challenges of devices for various use cases. Edge computing application primarily focuses on the classification of health data involving vital sign monitoring and fall detection. Other low-latency applications perform specific symptom monitoring for diseases, such as gait abnormalities in Parkinson's disease patients. We also present our exhaustive review on edge computing data operations that include transmission, encryption, authentication, classification, reduction, and prediction. Even with these advantages, edge computing has some associated challenges, including requirements for sophisticated privacy and data reduction methods to allow comparable performance to their Cloud-based counterparts, but with lower computational complexity. Future research directions in edge computing for health care have been identified to offer a higher quality of life for users if addressed.

Title: The dual CUSUM charts with auxiliary information for process mean Abstract: The CUSUM chart is usually designed based on a known shift size when monitoring the process mean. In practice, shift size is rarely known but it can be assumed to vary within an interval. With such a range of shift size, the dual CUSUM (DCUSUM) chart provides more sensitivity than the CUSUM chart. In this paper, auxiliary information based (AIB) DCUSUM and dual Crosier CUSUM (DCCUSUM) charts with and without fast initial response (FIR) features are proposed for efficiently monitoring the infrequent changes in the process mean, named the AIB-DCUSUM and AIB-DCCUSUM charts, respectively. Monte Carlo simulations are used to compute the run length characteristics of these control charts. The AIB-DCUSUM and AIB-DCCUSUM charts with and without FIR features are compared with the DCUSUM, DCCUSUM and AIB-CUSUM charts in terms of the integral relative average run length. It is found that the proposed charts show better performance when detecting mean shifts in different ranges. A simulated dataset is considered to illustrate the implementation of existing and proposed charts.

Title: A Piecewise Model for Two-phase Growth Phenomena Abstract: In this paper, we introduce a nonlinear model which can be used as a regression model for modeling phenomena requiring a two-phase growth curves. The proposed model is defined as a continuous piecewise combination of two exponential functions with an unknown break point. Numerical results on real and simulated data sets are presented to illustrate the applicability and the utility of the proposed model on growth data.

Title: A comprehensive review on recent advancements in routing protocols for flying ad hoc networks Abstract: Flying ad hoc networks (FANETs) are a particular type of ad hoc networks, which constitute flying nodes communicating with each other in ad hoc mode. Routing among FANETs has been a subject of interest due to the inherent challenge of handling dynamicity exhibited by the nodes. This survey presents a comprehensive review on recently developed routing protocols for FANETs. In the initial part of this paper, a basic overview of FANETs, their differences with respect to mobile ad hoc networks and vehicular ad hoc networks, evolution of FANET routing protocols, basic categorization of routing protocols, characteristics and design considerations for FANETs, has been presented. Then, an in-depth study of recently developed FANET routing protocols has been carried out by categorically differentiating them based on their routing mechanism, highlighting their features, strengths, weaknesses, and Quality of Service (QoS) parameters used for testing, in order to identify their suitability for routing data among flying nodes moving at high speeds. Furthermore, a feature comparison of all the discussed protocols has been presented in a summarized form. Lately, an analysis of mobility models, QoS parameters, simulators, and test beds used in the literature has been discussed in order to reflect their effectiveness when used for FANETs. Finally, open issues and future directions related to FANET routing protocols have been mentioned so as to direct the future research in this area for further advancements.

Title: An automatic short-answer grading model for semi-open-ended questions Abstract: Automatic short-answer grading has been studied for more than a decade. The technique has been used for implementing auto assessment as well as building the assessor module for intelligent tutoring systems. Many early works automatically grade mainly based on the similarity between a student answer and the reference answer to the question. This method performs well for closed-ended questions that have single or very limited numbers of correct answers. However, some short-answer questions ask students to express their own thoughts based on various facts; hence, they have no reference answers. Such questions are called semi-open-ended short-answer questions. Questions of this type often appear in reading comprehension assessments. In this paper, we developed an automatic semi-open-ended short-answer grading model that integrates both domain-general and domain-specific information. The model also utilizes a long-short-term-memory recurrent neural network to learn the representation in the classifier so that word sequence information is considered. In experiments on 7 reading comprehension questions and over 16,000 short-answer samples, our proposed automatic grading model demonstrates its advantage over existing models.

Title: Composite indicators in experimental psychology. An example with the semantic space of taste and shape stimuli stimuli Abstract: While composite indicators have been used in psychological sciences, they have mostly been confined within the area of social psychology for the assessment of specific issues in quality of life and health. In the area of experimental psychology composite indicators are far less familiar. This paper has the goal of presenting a procedure to design composite indicators that can be used for the analysis of experimental stimuli. An assessment of the extent to which matched taste and shape stimuli share a common semantic space shows the practical usefulness of the procedure. The robustness of the results was studied using uncertainty analysis.

Title: Locally weighted classifiers for detection of neighbor discovery protocol distributed denial-of-service and replayed attacks Abstract: The Internet of things requires more internet protocol (IP) addresses than IP version 4 (IPv4) can offer. To solve this problem, IP version 6 (IPv6) was developed to expand the availability of address spaces. Moreover, it supports hierarchical address allocation methods, which can facilitate route aggregation, thus limiting expansion of routing tables. An important feature of the IPv6 suites is the neighbor discovery protocol (NDP), which is geared towards substitution of the address resolution protocol in router discovery and function redirection in IPv4. However, NDP is vulnerable to denial-of-service (DoS) attacks. In this contribution, we present a novel detection method for distributed DoS (DDoS) attacks, launched using NDP in IPv6. The proposed system uses flow-based network representation, instead of a packet-based one. It exploits the advantages of locally weighted learning techniques, with three different machine learning models as its base learners. Simulation studies demonstrate that the intrusion detection method does not suffer from overfitting issues and offers lower computation costs and complexity, while exhibiting high accuracy rates. In summary, the proposed system uses six features, extracted from our bespoke dataset and is capable of detecting DDoS attacks with 99% accuracy and replayed attacks with an accuracy of 91.17%, offering a marked improvement in detection performance over state-of-the-art approaches.

Title: Towards a trusted unmanned aerial system using blockchain for the protection of critical infrastructure Abstract: With the exponential growth in the number of vital infrastructures such as nuclear plants and transport and distribution networks, these systems have become more susceptible to coordinated cyberattacks. One of the effective approaches used to strengthen the security of these infrastructures is the use of unmanned aerial vehicles (UAVs) for surveillance and data collection. However, UAVs themselves are prone to attacks on their collected sensor data. Recently, blockchain (BC) has been proposed as a revolutionary technology that can be integrated within Internet of things (IoT) to provide a desired level of security and privacy. However, the integration of BC within IoT networks, where UAV's sensors constitute a major component, is extremely challenging. The major contribution of this study is twofold:(1) survey the security issues for UAV's collected sensor data, define the security requirements for such systems, and identify ways to address them; and (2) propose a novel BC-based solution to ensure the security of and the trust between the UAVs and their relevant ground control stations. Our implementation results and analysis show that using UAVs as means for protecting critical infrastructure is greatly enhanced through the utilization of trusted BC-based unmanned aerial systems.

Title: Identification of influential observation in linear structural relationship model with known slope Abstract: A number of identification techniques are available in the literature to detect influential observations in linear regression models. However, the issue of the identification of influential observations in errors-in-variable models is still not very explored. In this paper we propose a new method for the identification of influential observations based on the COVRATIO statistic when the slope parameter is known. We determine the cut off point for this model on the basis of Monte Carlo simulation study and show that this cut off point performs well in the identification of influential observation in linear structural relationship model with known slope parameter. Finally, we present a real world example which also supports the findings obtained by the simulations earlier.

Title: Implementation of modified OLSR protocol in AANETs for UDP and TCP environment Abstract: Airborne Ad-hoc Networks (AANETs) are becoming highly popular nowadays. Due to the fast movements of aircraft, there are frequent topology updates which result into a link break between two communicating aircraft. The routing overhead increases multiple times in search of a new route continuously. To deal with this issue of increased overhead, this paper presents a new routing scheme named as Airborne-OLSR (AOLSR). The proposed scheme provides more optimization of Multi-point Relay (MPR) selection criteria used in existing Optimized Link State Routing (OLSR) protocol. This decreases the amount of overhead by selecting the MPR either on the right side or the left side of the source node which depends on the location of the node to which data is to be sent. The decrease in overhead will also result into more bandwidth availability which decreases the possibility of link break. The proposed routing scheme has been compared with the existing OLSR protocol for UDP and TCP environment with varying node speed in 3-D Gauss Markov Mobility model using network simulator-3(ns-3). The simulation analysis shows that the proposed scheme is better in terms of packet delivery ratio, End-to-End delay, routing overhead and throughput as compared to OLSR. We have also compared the AOLSR with previously developed AANET specific protocols like automatic dependent surveillance-broadcast system aided geographic routing protocol(A-GR) and geographic routing protocol for aircraft ad hoc network (GRAA).

Title: Energy-efficient system-on-chip reconfigurable architecture design for sum of absolute difference computation in motion estimation process of H.265/HEVC video encoding Abstract: Motion estimation is the important and computationally intensive part of any video encoding. The objective of this paper is to design and analyze the coarse and fine reconfiguration of processing element-based hardware design for block matching-based motion estimation in H.265/HEVC video processing. Sum of absolute difference (SAD) is the commonly used criteria for block matching in the motion estimation process. User input is taken as the parameter for coarse reconfiguration, and the threshold value in SAD is considered for the fine reconfiguration. Processing elements are the hardware units designed for performing SAD calculation. In this work, the hardware for block-based motion estimation for H.265/HEVC encoder is designed using multiple processing elements, which will calculate the SAD values. A system-on-chip architecture is implemented and verified the optimization in terms of power and area utilization by reconfiguring the architecture dynamically based on the input video quality requirement.

Title: Optimized Decision tree rules using divergence based grey wolf optimization for big data classification in health care Abstract: Most of the organizations are mainly focusing on large datasets for automatic mining of necessary information from big medical data. The major issue of the big medical data is about its complex data sets and volume, which is gradually increasing. This paper intends to propose a big data classification model (heart disease) in health care, which includes certain phases or steps. The steps are as follows: (1) Map-reduce framework (2) support vector machine (SVM) (3) optimized decision tree classifier (DT). Initially, the big data is supplied as the input to the MapReduce Framework, where it reduces the data content through some major operations. This framework uses the principle component analysis to reduce the dimensions of data. The reduced data is subjected to SVM, where it outputs the classes. The output data from SVM is processed with a new contribution called ‘Data transformation’ that paves way for optimal rule generation in decision tree classifier. The advanced optimization concept is involved in this process to optimize the weight and integer in data transformation. This paper introduces a new algorithm namely divergence based grey wolf optimization (DGWO). Finally, the transformed data is subjected to DT, where the classification takes place. The proposed DGWO model is compared over other conventional methods like firefly algorithm, artificial bee colony algorithm, particle swarm optimization algorithm, genetic algorithm and grey wolf optimizer algorithms.

Title: Designing and implementation of energy-efficient wireless photovoltaic monitoring system Abstract: Photovoltaic (PV) cell usually shows unpredictable results due to abrupt change in environment. Environmental parameters such as temperature, dust factor, irradiance, and air mass have direct influence on the output of PV cells. Monitoring PV and its related environmental parameters is very important for understanding and analysis of PV in practical conditions. The PV parameters have been found useful in many ways like reliability analysis, life cycle costing, returns on investment calculation, loss of load probability, and optimal designing of PV power supplies. Currently, designed PV monitoring systems (PVMS) are reported to be expensive, complex, and are limited in their applications, out of which only few are wireless. In this article, we design and implement a simple cost-effective and wireless PV monitoring solution. The design PVMS is based on microcontrollers, light-dependent resistors, current sensor, analog-to-digital converter, solid state relays, ZigBee (XBee) wireless microcontroller, and LabVIEW-based data logging software. The designed PVMS performs the real-time analysis and data logging of open circuit voltage (V-oc), short circuit current (J(sc)), ambient temperature (T-a), cell temperature (T-cell), and maximum power point. In addition, real-time I-V and P-V characterization of PV module is another valuable feature embedded in the designed PVMS.

Title: Probability-Based Online Algorithm for Switch Operation of Energy Efficient Data Center Abstract: The huge amount of energy consumed by the data centers around the world every year motivates the cloud service providers to operate data centers in a more energy efficient way. A promising solution is to turn off the idle servers, which, however, may be turned on later, incurring a significant startup cost. The problem turns to dynamically provisioning the workload, and cutting down the energy cos...

Title: TARNN: Task-aware autonomic resource management using neural networks in cloud environment Abstract: Resource management is one of the major issue in cloud computing for IaaS. Among several resource management problems allocation, provisioning, and requirement mapping directly affects the performance of cloud. Resource allocation signifies assignment of available resources to different workloads in an economically optimal manner. Precise and accurate allocation is required to maximize the usage of resources. Current method of task allocation do not take previously acquired knowledge, type of the tasks, and the QoS parameters altogether into account in the allocation phase, and it has not been trained for different set of tasks. Furthermore, the self-optimization of the autonomous system fails to address the task type and identify the relationship between tasks and the resource demands along with its requirements. Important aspects like task management and resource utilization are the primary factors to consider for such a characteristic. This paper will present a novel autonomic resource management framework named task-aware autonomic resource allocation strategy using neural networks (TARNN), which aims to use knowledge about the behavior of the task over an extended period of time and use this knowledge to allocate resources when a similar task is submitted in future by the user. To effectively do the allocation, a neural network-based approach is adopted to classify the tasks appropriately based on the task parameters, task type, and QoS parameters and allocate the resource optimally for a new task autonomously, without the intervention of the cloud provider. Moreover, to identify and improve the relationship of the tasks with the resources in the context of scheduling, we have proposed a novel modified Particle Swarm Optimization (m-PSO) algorithm to schedule the tasks to resources based on resource demands. In TARNN, we have separated the collected synthetic dataset into 60-40 ratio for training and testing purposes. We found that the neural network-based approach provides almost 80% accurate classification w.r.t. task type and QoS parameters. We have also compared our results with support vector machine (SVM) and got 69% accuracy. Since the tasks are classified appropriately, the occurrence of resource reconfiguration and VM migration is drastically reduced. Hence, our system provides better allocation of resources and schedules the tasks appropriately to the resources, thereby improving the performance of the cloud.

Title: On properties of empirical best predictors Abstract: The class of Empirical Best Predictors (EBP) is widely used in survey sampling and small area estimation due to possibility of prediction of any population or subpopulation parameters. Their good stochastic properties requires that a set of relatively strong assumptions should be met. We present results of simulation studies where properties of the predictor are widely studied including different cases of model misspecification. A fast algorithm of computing EBP is also applied.