Towards explanation of DNN-based prediction with guided feature inversion

Mengnan Du, Ninghao Liu, Qingquan Song, Xia Hu

Research output: Chapter in Book/Report/Conference proceedingConference contribution

79 Scopus citations

Abstract

While deep neural networks (DNN) have become an effective computational tool, the prediction results are often criticized by the lack of interpretability, which is essential in many real-world applications such as health informatics. Existing attempts based on local interpretations aim to identify relevant features contributing the most to the prediction of DNN by monitoring the neighborhood of a given input. They usually simply ignore the intermediate layers of the DNN that might contain rich information for interpretation. To bridge the gap, in this paper, we propose to investigate a guided feature inversion framework for taking advantage of the deep architectures towards effective interpretation. The proposed framework not only determines the contribution of each feature in the input but also provides insights into the decision-making process of DNN models. By further interacting with the neuron of the target category at the output layer of the DNN, we enforce the interpretation result to be class-discriminative. We apply the proposed interpretation model to different CNN architectures to provide explanations for image data and conduct extensive experiments on ImageNet and PASCAL VOC07 datasets. The interpretation results demonstrate the effectiveness of our proposed framework in providing class-discriminative interpretation for DNN-based prediction.

Original languageEnglish (US)
Title of host publicationKDD 2018 - Proceedings of the 24th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining
PublisherAssociation for Computing Machinery
Pages1358-1367
Number of pages10
ISBN (Print)9781450355520
DOIs
StatePublished - Jul 19 2018
Externally publishedYes
Event24th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, KDD 2018 - London, United Kingdom
Duration: Aug 19 2018Aug 23 2018

Publication series

NameProceedings of the ACM SIGKDD International Conference on Knowledge Discovery and Data Mining

Conference

Conference24th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, KDD 2018
Country/TerritoryUnited Kingdom
CityLondon
Period8/19/188/23/18

All Science Journal Classification (ASJC) codes

  • Software
  • Information Systems

Keywords

  • Deep learning
  • Guided feature inversion
  • Intermediate layers
  • Machine learning interpretation

Fingerprint

Dive into the research topics of 'Towards explanation of DNN-based prediction with guided feature inversion'. Together they form a unique fingerprint.

Cite this