Learning Granger Causality from Instance-wise Self-attentive Hawkes Processes

Dongxia Wu, Tsuyoshi Ide, Georgios Kollias, Jiri Navratil, Aurelie Lozano, Naoki Abe, Yian Ma, Rose Yu
Proceedings of The 27th International Conference on Artificial Intelligence and Statistics, PMLR 238:415-423, 2024.

Abstract

We address the problem of learning Granger causality from asynchronous, interdependent, multi-type event sequences. In particular, we are interested in discovering instance-level causal structures in an unsupervised manner. Instance-level causality identifies causal relationships among individual events, providing more fine-grained information for decision-making. Existing work in the literature either requires strong assumptions, such as linearity in the intensity function, or heuristically defined model parameters that do not necessarily meet the requirements of Granger causality. We propose Instance-wise Self-Attentive Hawkes Processes (ISAHP), a novel deep learning framework that can directly infer the Granger causality at the event instance level. ISAHP is the first neural point process model that meets the requirements of Granger causality. It leverages the self-attention mechanism of the transformer to align with the principles of Granger causality. We empirically demonstrate that ISAHP is capable of discovering complex instance-level causal structures that cannot be handled by classical models. We also show that ISAHP achieves state-of-the-art performance in proxy tasks involving type-level causal discovery and instance-level event type prediction.

Cite this Paper

BibTeX
@InProceedings{pmlr-v238-wu24a, title = {Learning {G}ranger Causality from Instance-wise Self-attentive {H}awkes Processes}, author = {Wu, Dongxia and Ide, Tsuyoshi and Kollias, Georgios and Navratil, Jiri and Lozano, Aurelie and Abe, Naoki and Ma, Yian and Yu, Rose}, booktitle = {Proceedings of The 27th International Conference on Artificial Intelligence and Statistics}, pages = {415--423}, year = {2024}, editor = {Dasgupta, Sanjoy and Mandt, Stephan and Li, Yingzhen}, volume = {238}, series = {Proceedings of Machine Learning Research}, month = {02--04 May}, publisher = {PMLR}, pdf = {https://proceedings.mlr.press/v238/wu24a/wu24a.pdf}, url = {https://proceedings.mlr.press/v238/wu24a.html}, abstract = {We address the problem of learning Granger causality from asynchronous, interdependent, multi-type event sequences. In particular, we are interested in discovering instance-level causal structures in an unsupervised manner. Instance-level causality identifies causal relationships among individual events, providing more fine-grained information for decision-making. Existing work in the literature either requires strong assumptions, such as linearity in the intensity function, or heuristically defined model parameters that do not necessarily meet the requirements of Granger causality. We propose Instance-wise Self-Attentive Hawkes Processes (ISAHP), a novel deep learning framework that can directly infer the Granger causality at the event instance level. ISAHP is the first neural point process model that meets the requirements of Granger causality. It leverages the self-attention mechanism of the transformer to align with the principles of Granger causality. We empirically demonstrate that ISAHP is capable of discovering complex instance-level causal structures that cannot be handled by classical models. We also show that ISAHP achieves state-of-the-art performance in proxy tasks involving type-level causal discovery and instance-level event type prediction.} }
Endnote
%0 Conference Paper %T Learning Granger Causality from Instance-wise Self-attentive Hawkes Processes %A Dongxia Wu %A Tsuyoshi Ide %A Georgios Kollias %A Jiri Navratil %A Aurelie Lozano %A Naoki Abe %A Yian Ma %A Rose Yu %B Proceedings of The 27th International Conference on Artificial Intelligence and Statistics %C Proceedings of Machine Learning Research %D 2024 %E Sanjoy Dasgupta %E Stephan Mandt %E Yingzhen Li %F pmlr-v238-wu24a %I PMLR %P 415--423 %U https://proceedings.mlr.press/v238/wu24a.html %V 238 %X We address the problem of learning Granger causality from asynchronous, interdependent, multi-type event sequences. In particular, we are interested in discovering instance-level causal structures in an unsupervised manner. Instance-level causality identifies causal relationships among individual events, providing more fine-grained information for decision-making. Existing work in the literature either requires strong assumptions, such as linearity in the intensity function, or heuristically defined model parameters that do not necessarily meet the requirements of Granger causality. We propose Instance-wise Self-Attentive Hawkes Processes (ISAHP), a novel deep learning framework that can directly infer the Granger causality at the event instance level. ISAHP is the first neural point process model that meets the requirements of Granger causality. It leverages the self-attention mechanism of the transformer to align with the principles of Granger causality. We empirically demonstrate that ISAHP is capable of discovering complex instance-level causal structures that cannot be handled by classical models. We also show that ISAHP achieves state-of-the-art performance in proxy tasks involving type-level causal discovery and instance-level event type prediction.
APA
Wu, D., Ide, T., Kollias, G., Navratil, J., Lozano, A., Abe, N., Ma, Y. & Yu, R.. (2024). Learning Granger Causality from Instance-wise Self-attentive Hawkes Processes. Proceedings of The 27th International Conference on Artificial Intelligence and Statistics, in Proceedings of Machine Learning Research 238:415-423 Available from https://proceedings.mlr.press/v238/wu24a.html.

Related Material