Towards Interaction Detection Using Topological Analysis on Neural Networks

TL;DR

This paper introduces a novel topological approach using persistent homology to detect feature interactions in neural networks, providing a new measure of interaction strength and an efficient detection algorithm validated on various datasets.

Contribution

It proposes a new topological measure for interaction strength and a Persistence Interaction detection (PID) algorithm, advancing the state-of-the-art in neural network interaction detection.

Findings

PID outperforms existing methods on synthetic datasets.

The topological measure effectively captures interaction strength.

The approach is validated on real-world datasets.

Abstract

Detecting statistical interactions between input features is a crucial and challenging task. Recent advances demonstrate that it is possible to extract learned interactions from trained neural networks. It has also been observed that, in neural networks, any interacting features must follow a strongly weighted connection to common hidden units. Motivated by the observation, in this paper, we propose to investigate the interaction detection problem from a novel topological perspective by analyzing the connectivity in neural networks. Specially, we propose a new measure for quantifying interaction strength, based upon the well-received theory of persistent homology. Based on this measure, a Persistence Interaction detection~(PID) algorithm is developed to efficiently detect interactions. Our proposed algorithm is evaluated across a number of interaction detection tasks on several synthetic and real world datasets with different hyperparameters. Experimental results validate that the PID algorithm outperforms the state-of-the-art baselines.