Learning Time-Aware Causal Representation for Model Generalization in Evolving Domains

Zhuo He, Shuang Li, Wenze Song, Longhui Yuan, Jian Liang, Han Li, Kun Gai
Proceedings of the 42nd International Conference on Machine Learning, PMLR 267:22523-22547, 2025.

Abstract

Endowing deep models with the ability to generalize in dynamic scenarios is of vital significance for real-world deployment, given the continuous and complex changes in data distribution. Recently, evolving domain generalization (EDG) has emerged to address distribution shifts over time, aiming to capture evolving patterns for improved model generalization. However, existing EDG methods may suffer from spurious correlations by modeling only the dependence between data and targets across domains, creating a shortcut between task-irrelevant factors and the target, which hinders generalization. To this end, we design a time-aware structural causal model (SCM) that incorporates dynamic causal factors and the causal mechanism drifts, and propose Static-DYNamic Causal Representation Learning (SYNC), an approach that effectively learns time-aware causal representations. Specifically, it integrates specially designed information-theoretic objectives into a sequential VAE framework which captures evolving patterns, and produces the desired representations by preserving intra-class compactness of causal factors both across and within domains. Moreover, we theoretically show that our method can yield the optimal causal predictor for each time domain. Results on both synthetic and real-world datasets exhibit that SYNC can achieve superior temporal generalization performance.

Cite this Paper

BibTeX
@InProceedings{pmlr-v267-he25j, title = {Learning Time-Aware Causal Representation for Model Generalization in Evolving Domains}, author = {He, Zhuo and Li, Shuang and Song, Wenze and Yuan, Longhui and Liang, Jian and Li, Han and Gai, Kun}, booktitle = {Proceedings of the 42nd International Conference on Machine Learning}, pages = {22523--22547}, year = {2025}, editor = {Singh, Aarti and Fazel, Maryam and Hsu, Daniel and Lacoste-Julien, Simon and Berkenkamp, Felix and Maharaj, Tegan and Wagstaff, Kiri and Zhu, Jerry}, volume = {267}, series = {Proceedings of Machine Learning Research}, month = {13--19 Jul}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v267/main/assets/he25j/he25j.pdf}, url = {https://proceedings.mlr.press/v267/he25j.html}, abstract = {Endowing deep models with the ability to generalize in dynamic scenarios is of vital significance for real-world deployment, given the continuous and complex changes in data distribution. Recently, evolving domain generalization (EDG) has emerged to address distribution shifts over time, aiming to capture evolving patterns for improved model generalization. However, existing EDG methods may suffer from spurious correlations by modeling only the dependence between data and targets across domains, creating a shortcut between task-irrelevant factors and the target, which hinders generalization. To this end, we design a time-aware structural causal model (SCM) that incorporates dynamic causal factors and the causal mechanism drifts, and propose Static-DYNamic Causal Representation Learning (SYNC), an approach that effectively learns time-aware causal representations. Specifically, it integrates specially designed information-theoretic objectives into a sequential VAE framework which captures evolving patterns, and produces the desired representations by preserving intra-class compactness of causal factors both across and within domains. Moreover, we theoretically show that our method can yield the optimal causal predictor for each time domain. Results on both synthetic and real-world datasets exhibit that SYNC can achieve superior temporal generalization performance.} }
Endnote
%0 Conference Paper %T Learning Time-Aware Causal Representation for Model Generalization in Evolving Domains %A Zhuo He %A Shuang Li %A Wenze Song %A Longhui Yuan %A Jian Liang %A Han Li %A Kun Gai %B Proceedings of the 42nd International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2025 %E Aarti Singh %E Maryam Fazel %E Daniel Hsu %E Simon Lacoste-Julien %E Felix Berkenkamp %E Tegan Maharaj %E Kiri Wagstaff %E Jerry Zhu %F pmlr-v267-he25j %I PMLR %P 22523--22547 %U https://proceedings.mlr.press/v267/he25j.html %V 267 %X Endowing deep models with the ability to generalize in dynamic scenarios is of vital significance for real-world deployment, given the continuous and complex changes in data distribution. Recently, evolving domain generalization (EDG) has emerged to address distribution shifts over time, aiming to capture evolving patterns for improved model generalization. However, existing EDG methods may suffer from spurious correlations by modeling only the dependence between data and targets across domains, creating a shortcut between task-irrelevant factors and the target, which hinders generalization. To this end, we design a time-aware structural causal model (SCM) that incorporates dynamic causal factors and the causal mechanism drifts, and propose Static-DYNamic Causal Representation Learning (SYNC), an approach that effectively learns time-aware causal representations. Specifically, it integrates specially designed information-theoretic objectives into a sequential VAE framework which captures evolving patterns, and produces the desired representations by preserving intra-class compactness of causal factors both across and within domains. Moreover, we theoretically show that our method can yield the optimal causal predictor for each time domain. Results on both synthetic and real-world datasets exhibit that SYNC can achieve superior temporal generalization performance.
APA
He, Z., Li, S., Song, W., Yuan, L., Liang, J., Li, H. & Gai, K.. (2025). Learning Time-Aware Causal Representation for Model Generalization in Evolving Domains. Proceedings of the 42nd International Conference on Machine Learning, in Proceedings of Machine Learning Research 267:22523-22547 Available from https://proceedings.mlr.press/v267/he25j.html.

Related Material