This repository contains a software implementation of the methods described in:

Predictive Multiplicity in Classification
Charlie Marx, Flavio Calmon, Berk Ustun

Installation

Requirements

• Python 3
• CPLEX

CPLEX

CPLEX is cross-platform commercial optimization tool with a Python API. It is free for students and faculty at accredited institutions. To get CPLEX:

1. Register for IBM OnTheHub
2. Download the IBM ILOG CPLEX Optimization Studio from the software catalog
3. Install the CPLEX Optimization Studio.
4. Setup the CPLEX Python API as described here.

If you have problems installing CPLEX, check the CPLEX user manual or the CPLEX forums.

Usage

Example usage can be found in the example.py file.

Development

The code in this repository is currently under active development, and may therefore change substantially with each commit.

Reference

If you use our code, please cite our paper.

@misc{marx2019predictive,
    title={Predictive Multiplicity in Classification},
    author={Charles T. Marx and Flavio du Pin Calmon and Berk Ustun},
    year={2019},
    eprint={1909.06677},
    archivePrefix={arXiv},
    primaryClass={cs.LG}
}