Open Peer Review. Open Publishing. Open Access. Open Discussion. Open Directory. Open Recommendations. Open API. Open Source.
Detecting Statistical Interactions from Neural Network Weights
Michael Tsang, Dehua Cheng, Yan Liu
Feb 15, 2018 (modified: Feb 28, 2018)ICLR 2018 Conference Blind Submissionreaders: everyoneShow Bibtex
Abstract:Interpreting neural networks is a crucial and challenging task in machine learning. In this paper, we develop a novel framework for detecting statistical interactions captured by a feedforward multilayer neural network by directly interpreting its learned weights. Depending on the desired interactions, our method can achieve significantly better or similar interaction detection performance compared to the state-of-the-art without searching an exponential solution space of possible interactions. We obtain this accuracy and efficiency by observing that interactions between input features are created by the non-additive effect of nonlinear activation functions, and that interacting paths are encoded in weight matrices. We demonstrate the performance of our method and the importance of discovered interactions via experimental results on both synthetic datasets and real-world application datasets.
TL;DR:We detect statistical interactions captured by a feedforward multilayer neural network by directly interpreting its learned weights.
Keywords:statistical interaction detection, multilayer perceptron, generalized additive model
Enter your feedback below and we'll get back to you as soon as possible.