Learning Perturbations to Explain Time Series Predictions

Joseph Enguehard
Proceedings of the 40th International Conference on Machine Learning, PMLR 202:9329-9342, 2023.

Abstract

Explaining predictions based on multivariate time series data carries the additional difficulty of handling not only multiple features, but also time dependencies. It matters not only what happened, but also when, and the same feature could have a very different impact on a prediction depending on this time information. Previous work has used perturbation-based saliency methods to tackle this issue, perturbing an input using a trainable mask to discover which features at which times are driving the predictions. However these methods introduce fixed perturbations, inspired from similar methods on static data, while there seems to be little motivation to do so on temporal data. In this work, we aim to explain predictions by learning not only masks, but also associated perturbations. We empirically show that learning these perturbations significantly improves the quality of these explanations on time series data.

Cite this Paper

BibTeX
@InProceedings{pmlr-v202-enguehard23a, title = {Learning Perturbations to Explain Time Series Predictions}, author = {Enguehard, Joseph}, booktitle = {Proceedings of the 40th International Conference on Machine Learning}, pages = {9329--9342}, year = {2023}, editor = {Krause, Andreas and Brunskill, Emma and Cho, Kyunghyun and Engelhardt, Barbara and Sabato, Sivan and Scarlett, Jonathan}, volume = {202}, series = {Proceedings of Machine Learning Research}, month = {23--29 Jul}, publisher = {PMLR}, pdf = {https://proceedings.mlr.press/v202/enguehard23a/enguehard23a.pdf}, url = {https://proceedings.mlr.press/v202/enguehard23a.html}, abstract = {Explaining predictions based on multivariate time series data carries the additional difficulty of handling not only multiple features, but also time dependencies. It matters not only what happened, but also when, and the same feature could have a very different impact on a prediction depending on this time information. Previous work has used perturbation-based saliency methods to tackle this issue, perturbing an input using a trainable mask to discover which features at which times are driving the predictions. However these methods introduce fixed perturbations, inspired from similar methods on static data, while there seems to be little motivation to do so on temporal data. In this work, we aim to explain predictions by learning not only masks, but also associated perturbations. We empirically show that learning these perturbations significantly improves the quality of these explanations on time series data.} }
Endnote
%0 Conference Paper %T Learning Perturbations to Explain Time Series Predictions %A Joseph Enguehard %B Proceedings of the 40th International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2023 %E Andreas Krause %E Emma Brunskill %E Kyunghyun Cho %E Barbara Engelhardt %E Sivan Sabato %E Jonathan Scarlett %F pmlr-v202-enguehard23a %I PMLR %P 9329--9342 %U https://proceedings.mlr.press/v202/enguehard23a.html %V 202 %X Explaining predictions based on multivariate time series data carries the additional difficulty of handling not only multiple features, but also time dependencies. It matters not only what happened, but also when, and the same feature could have a very different impact on a prediction depending on this time information. Previous work has used perturbation-based saliency methods to tackle this issue, perturbing an input using a trainable mask to discover which features at which times are driving the predictions. However these methods introduce fixed perturbations, inspired from similar methods on static data, while there seems to be little motivation to do so on temporal data. In this work, we aim to explain predictions by learning not only masks, but also associated perturbations. We empirically show that learning these perturbations significantly improves the quality of these explanations on time series data.
APA
Enguehard, J.. (2023). Learning Perturbations to Explain Time Series Predictions. Proceedings of the 40th International Conference on Machine Learning, in Proceedings of Machine Learning Research 202:9329-9342 Available from https://proceedings.mlr.press/v202/enguehard23a.html.

Related Material