Airlines: Subsampled Classification Dataset with 10 Classes and 100 Features

Airlines is a subsampled classification dataset from OpenML, derived from dataset ID 1169. It was generated by Eddie Bergman using a script to create a version with a maximum of 2000 rows, 100 columns, and 10 target classes. The subsampling process used a random seed of 0 and applied stratified sampling.

Use Cases

• Benchmarking classification model performance based on the described 10 target classes.
• Feature selection experiments based on the described 100 sampled columns.
• Testing the impact of stratified sampling on model evaluation using the described generation method.

Strengths

• Dataset creation parameters are explicitly documented, including a random seed of 0, a maximum of 2000 rows, 100 columns, and 10 classes.
• The generation code indicates stratified sampling was used, which can help preserve class distribution.

Limitations

• The original dataset's row count, column names, and sample data are unknown, limiting content assessment.
• Column-level documentation is absent; field semantics must be inferred after download.
• Last update date is unknown; freshness unverified.

Provenance

Source

OpenML, derived from dataset 'airlines' (ID 1169).

Collection Method

Algorithmic subsampling of an original dataset using a specified Python script.