Datasets:
metadata
language:
- en
tags:
- madelon
- tabular_classification
- UCI
pretty_name: Madelon
size_categories:
- 1K<n<10K
task_categories:
- tabular-classification
configs:
- Madelon
license: cc
Annealing
The Madelon dataset from the UCI ML repository. Artificial dataset with continuous input variables. Highly non-linear classification problem.
Configurations and tasks
Configuration | Task | Description |
---|---|---|
madelon | Binary classification |
Usage
from datasets import load_dataset
dataset = load_dataset("mstz/madelon")["train"]