Neural Additive Models: Interpretable Machine Learning with Neural Nets

This repository contains open-source code for the paper Neural Additive Models: Interpretable Machine Learning with Neural Nets.

Currently, we release the tf.keras.Model for NAM which can be simply plugged into any neural network training procedure. We also provide helpers for building a computation graph using NAM for classification/regression problems with tf.compat.v1. The nam_train.py file provides the example of a training script on a single dataset split.

Use ./run.sh test script to ensure that the setup is correct.

Dependencies

The code was tested under Ubuntu 16 and uses these packages:

• tensorflow>=1.15
• numpy>=1.15.2
• sklearn>=0.23
• pandas>=0.24
• absl-py

Citing

If you use this code in your research, please cite the following paper:

Agarwal, R., Frosst, N., Zhang, X., Caruana, R., & Hinton, G. E. (2020). Neural additive models: Interpretable machine learning with neural nets. arXiv preprint arXiv:2004.13912

  @article{agarwal2020neural,
    title={Neural additive models: Interpretable machine learning with neural nets},
    author={Agarwal, Rishabh and Frosst, Nicholas and Zhang, Xuezhou and
    Caruana, Rich and Hinton, Geoffrey E},
    journal={arXiv preprint arXiv:2004.13912},
    year={2020}
  }

---

Disclaimer about COMPAS dataset: It is important to note that developing a machine learning model to predict pre-trial detention has a number of important ethical considerations. You can learn more about these issues in the Partnership on AI Report on Algorithmic Risk Assessment Tools in the U.S. Criminal Justice System. The Partnership on AI is a multi-stakeholder organization -- of which Google is a member -- that creates guidelines around AI.

We’re using the COMPAS dataset only as an example of how to identify and remediate fairness concerns in data. This dataset is canonical in the algorithmic fairness literature.

Disclaimer: This is not an official Google product.