TY - GEN
T1 - An Operator Theoretic Approach for Analyzing Sequence Neural Networks
AU - Naiman, Ilan
AU - Azencot, Omri
N1 - Funding Information:
This research was partially supported by the Lynn and William Frankel Center of the Computer Science Department, Ben-Gurion University of the Negev, an ISF grant 668/21, an ISF equipment grant, and by the Israeli Council for Higher Education (CHE) via Data Science Research Center, Ben-Gurion University of the Negev, Israel.
Publisher Copyright:
Copyright © 2023, Association for the Advancement of Artificial Intelligence (www.aaai.org). All rights reserved.
PY - 2023/6/27
Y1 - 2023/6/27
N2 - Analyzing the inner mechanisms of deep neural networks is a fundamental task in machine learning. Existing work provides limited analysis or it depends on local theories, such as fixed-point analysis. In contrast, we propose to analyze trained neural networks using an operator theoretic approach which is rooted in Koopman theory, the Koopman Analysis of Neural Networks (KANN). Key to our method is the Koopman operator, which is a linear object that globally represents the dominant behavior of the network dynamics. The linearity of the Koopman operator facilitates analysis via its eigenvectors and eigenvalues. Our method reveals that the latter eigendecomposition holds semantic information related to the neural network inner workings. For instance, the eigenvectors highlight positive and negative n-grams in the sentiments analysis task; similarly, the eigenvectors capture the salient features of healthy heart beat signals in the ECG classification problem.
AB - Analyzing the inner mechanisms of deep neural networks is a fundamental task in machine learning. Existing work provides limited analysis or it depends on local theories, such as fixed-point analysis. In contrast, we propose to analyze trained neural networks using an operator theoretic approach which is rooted in Koopman theory, the Koopman Analysis of Neural Networks (KANN). Key to our method is the Koopman operator, which is a linear object that globally represents the dominant behavior of the network dynamics. The linearity of the Koopman operator facilitates analysis via its eigenvectors and eigenvalues. Our method reveals that the latter eigendecomposition holds semantic information related to the neural network inner workings. For instance, the eigenvectors highlight positive and negative n-grams in the sentiments analysis task; similarly, the eigenvectors capture the salient features of healthy heart beat signals in the ECG classification problem.
UR - http://www.scopus.com/inward/record.url?scp=85152279814&partnerID=8YFLogxK
M3 - Conference contribution
AN - SCOPUS:85152279814
T3 - Proceedings of the 37th AAAI Conference on Artificial Intelligence, AAAI 2023
SP - 9268
EP - 9276
BT - AAAI-23 Technical Tracks 8
A2 - Williams, Brian
A2 - Chen, Yiling
A2 - Neville, Jennifer
PB - AAAI press
T2 - 37th AAAI Conference on Artificial Intelligence, AAAI 2023
Y2 - 7 February 2023 through 14 February 2023
ER -