NODE-GAM: Neural Generalized Additive Model for Interpretable Deep Learning

Deployment of machine learning models in real high-risk settings (e.g. healthcare) often depends not only on model’s accuracy but also on its fairness, robustness and interpretability. Generalized Additive Models (GAMs) have a long history of use in these high-risk domains, but lack desirable features of deep learning such as differentiability and scalability. In this work, we propose a neural GAM (NODE-GAM) and neural GAM (NODE-GAM) that scale well to large datasets, while remaining interpretable and accurate. We show that our proposed models have comparable accuracy to other non-interpretable models, and outperform other GAMs on large datasets. We also show that our models are more accurate in self-supervised learning setting when access to labeled data is limited.

[1]  Yoshua Bengio,et al.  Show, Attend and Tell: Neural Image Caption Generation with Visual Attention , 2015, ICML.

[2]  Rich Caruana,et al.  Distill-and-Compare: Auditing Black-Box Models Using Transparent Model Distillation , 2017, AIES.

[3]  Jerry Ma,et al.  Quasi-hyperbolic momentum and Adam for deep learning , 2018, ICLR.

[4]  Peter Szolovits,et al.  MIMIC-III, a freely accessible critical care database , 2016, Scientific Data.

[5]  Rich Caruana,et al.  InterpretML: A Unified Framework for Machine Learning Interpretability , 2019, ArXiv.

[6]  S. Sapra,et al.  Generalized additive models in business and economics , 2013 .

[7]  Yan Liu,et al.  Neural Interaction Transparency (NIT): Disentangling Learned Interactions for Improved Interpretability , 2018, NeurIPS.

[8]  R. Tibshirani,et al.  Generalized Additive Models , 1986 .

[9]  Yuval Pinter,et al.  Attention is not not Explanation , 2019, EMNLP.

[10]  André F. T. Martins,et al.  Sparse Sequence-to-Sequence Models , 2019, ACL.

[11]  Sercan O. Arik,et al.  TabNet: Attentive Interpretable Tabular Learning , 2019, AAAI.

[12]  Johannes Gehrke,et al.  Intelligible Models for HealthCare: Predicting Pneumonia Risk and Hospital 30-day Readmission , 2015, KDD.

[13]  Stefan Hegselmann,et al.  An Evaluation of the Doctor-Interpretability of Generalized Additive Models with Interactions , 2020, MLHC.

[14]  R. Tibshirani,et al.  Generalized additive models for medical research , 1995, Statistical methods in medical research.

[15]  Harmanpreet Kaur,et al.  Interpreting Interpretability: Understanding Data Scientists' Use of Interpretability Tools for Machine Learning , 2020, CHI.

[16]  Farzali Izadi,et al.  Generalized Additive Models to Capture the Death Rates in Canada COVID-19 , 2020, Mathematics of Public Health.

[17]  Rich Caruana,et al.  Purifying Interaction Effects with the Functional ANOVA: An Efficient Algorithm for Recovering Identifiable Additive Models , 2020, AISTATS.

[18]  Andrew Gordon Wilson,et al.  Averaging Weights Leads to Wider Optima and Better Generalization , 2018, UAI.

[19]  Kaiming He,et al.  Accurate, Large Minibatch SGD: Training ImageNet in 1 Hour , 2017, ArXiv.

[20]  Daniel Servén,et al.  pyGAM: Generalized Additive Models in Python , 2018 .

[21]  Rich Caruana,et al.  How Interpretable and Trustworthy are GAMs? , 2020, KDD.

[22]  Sarah Tan,et al.  Learning Global Additive Explanations of Black-Box Models , 2019 .

[23]  Eric J. Pedersen,et al.  Hierarchical generalized additive models in ecology: an introduction with mgcv , 2019, PeerJ.

[24]  Albert Gordo,et al.  Learning Global Additive Explanations for Neural Nets Using Model Distillation , 2018 .

[25]  Geoffrey E. Hinton,et al.  Neural Additive Models: Interpretable Machine Learning with Neural Nets , 2020, NeurIPS.

[26]  Been Kim,et al.  Concept Bottleneck Models , 2020, ICML.

[27]  Johannes Gehrke,et al.  Accurate intelligible models with pairwise interactions , 2013, KDD.

[28]  Sergei Popov,et al.  Neural Oblivious Decision Ensembles for Deep Learning on Tabular Data , 2019, ICLR.

[29]  Yan Liu,et al.  Detecting Statistical Interactions from Neural Network Weights , 2017, ICLR.