Methods for interpreting and understanding deep neural networks

Montavon, Grégoire; Samek, Wojciech; Müller, Klaus‐Robert

doi:10.1016/j.dsp.2017.10.011

articleDigital Signal ProcessingOct 24, 2017HYBRID OA

Methods for interpreting and understanding deep neural networks

GMGrégoire Montavon WSWojciech Samek KMKlaus‐Robert Müller

Technische Universität Berlin · Fraunhofer Institute for Telecommunications, Heinrich Hertz Institute · +2 more institutions

Indexed inarxivcrossref

Abstract

This paper provides an entry point to the problem of interpreting a deep neural network model and explaining its predictions. It is based on a tutorial given at ICASSP 2017. As a tutorial paper, the set of methods covered here is not exhaustive, but sufficiently representative to discuss a number of questions in interpretability, technical challenges, and possible applications. The second part of the tutorial focuses on the recently proposed layer-wise relevance propagation (LRP) technique, for which we provide theory, recommendations, and tricks, to make most efficient use of it on real data.

Citation impact

2,686

total citations

FWCI: 158.36
Percentile: 100%
References: 122

Citations per year

Authors

3

Topics & keywords

Topics

Keywords

Computer science
Artificial neural network
Artificial intelligence
Deep neural networks

No related works found for this paper.