DyCVAE: Learning Dynamic Causal Factors for Non-stationary Series Domain Generalization (Student Abstract)

Authors

  • Weifeng Zhang University of Electronic Science and Technology of China, China
  • Zhiyuan Wang University of Electronic Science and Technology of China, China
  • Kunpeng Zhang University of Maryland, College Park, USA
  • Ting Zhong University of Electronic Science and Technology of China, China
  • Fan Zhou University of Electronic Science and Technology of China, China Kashi Institute of Electronics and Information Industry, China

DOI:

https://doi.org/10.1609/aaai.v37i13.27051

Keywords:

Domain Generalization, Non-stationary Time Series, Causal Representation

Abstract

Learning domain-invariant representations is a major task of out-of-distribution generalization. To address this issue, recent efforts have taken into accounting causality, aiming at learning the causal factors with regard to tasks. However, extending existing generalization methods for adapting non-stationary time series may be ineffective, because they fail to model the underlying causal factors due to temporal-domain shifts except for source-domain shifts, as pointed out by recent studies. To this end, we propose a novel model DyCVAE to learn dynamic causal factors. The results on synthetic and real datasets demonstrate the effectiveness of our proposed model for the task of generalization in time series domain.
AAAI-23 Proceedings Cover

Downloads

Published

2023-09-06

How to Cite

Zhang, W., Wang, Z., Zhang, K., Zhong, T., & Zhou, F. (2023). DyCVAE: Learning Dynamic Causal Factors for Non-stationary Series Domain Generalization (Student Abstract). Proceedings of the AAAI Conference on Artificial Intelligence, 37(13), 16382-16383. https://doi.org/10.1609/aaai.v37i13.27051

Issue

Vol. 37 No. 13: AAAI-23 Special Programs, IAAI-23, EAAI-23, Student Papers and Demonstrations

Section

AAAI Student Abstract and Poster Program