TY - GEN
T1 - Explanation of Multi-Label Neural Networks with Layer-Wise Relevance Propagation
AU - Bello, Marilyn
AU - Napoles, Gonzalo
AU - Vanhoof, Koen
AU - Garcia, Maria M.
AU - Bello, Rafael
N1 - Funding Information:
The authors would like to sincerely thank the doctors of the Hospital “Comandante Manuel Fajardo Rivero” in the city of Santa Clara, Cuba, who assisted us in the description and medical terminology associated with the case study under consideration. Likewise, in the validation of the results obtained as an expert in the field. This study is supported by the Special Research Fund of Hasselt University.
Publisher Copyright:
© 2022 IEEE.
PY - 2022
Y1 - 2022
N2 - Neural networks are considered a black-box model as their strength in modeling complex interactions makes its operation almost impossible to explain. Still, neural networks remain very interesting tools as they have shown promising performance in various classification tasks. Layer-wise relevance propagation is a technique that, based on a propagation approach, is able to explain the predictions obtained by a neural network. In this work, we propose four adaptations of this technique to operate on multi-label neural networks. The proposed methods provide new ways of distributing the relevance between the output layer and the preceding ones. The efficacy of these adaptations is demonstrated after an experimental study. The study is carried out based on existing evaluation criteria in the literature that measure the explanation's quality. These methods are applied to a case study in which a neural network is used to detect secondary coinfections in patients infected with SARS-CoV-2. Overall, the proposed methods provide a post-hoc interpretability stage of the results.
AB - Neural networks are considered a black-box model as their strength in modeling complex interactions makes its operation almost impossible to explain. Still, neural networks remain very interesting tools as they have shown promising performance in various classification tasks. Layer-wise relevance propagation is a technique that, based on a propagation approach, is able to explain the predictions obtained by a neural network. In this work, we propose four adaptations of this technique to operate on multi-label neural networks. The proposed methods provide new ways of distributing the relevance between the output layer and the preceding ones. The efficacy of these adaptations is demonstrated after an experimental study. The study is carried out based on existing evaluation criteria in the literature that measure the explanation's quality. These methods are applied to a case study in which a neural network is used to detect secondary coinfections in patients infected with SARS-CoV-2. Overall, the proposed methods provide a post-hoc interpretability stage of the results.
KW - Explanation
KW - Layer-wise Relevance Propagation
KW - Multi-label Scenarios
KW - Neural Networks
UR - http://www.scopus.com/inward/record.url?scp=85140759793&partnerID=8YFLogxK
U2 - 10.1109/IJCNN55064.2022.9892239
DO - 10.1109/IJCNN55064.2022.9892239
M3 - Conference contribution
AN - SCOPUS:85140759793
T3 - Proceedings of the International Joint Conference on Neural Networks
BT - 2022 International Joint Conference on Neural Networks, IJCNN 2022 - Proceedings
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 2022 International Joint Conference on Neural Networks, IJCNN 2022
Y2 - 18 July 2022 through 23 July 2022
ER -