Gradient-based Explanations for Deep Learning Survival Models

Sophie Hanna Langbein, Niklas Koenen, Marvin N. Wright
Proceedings of the 42nd International Conference on Machine Learning, PMLR 267:32492-32522, 2025.

Abstract

Deep learning survival models often outperform classical methods in time-to-event predictions, particularly in personalized medicine, but their "black box" nature hinders broader adoption. We propose a framework for gradient-based explanation methods tailored to survival neural networks, extending their use beyond regression and classification. We analyze the implications of their theoretical assumptions for time-dependent explanations in the survival setting and propose effective visualizations incorporating the temporal dimension. Experiments on synthetic data show that gradient-based methods capture the magnitude and direction of local and global feature effects, including time dependencies. We introduce GradSHAP(t), a gradient-based counterpart to SurvSHAP(t), which outperforms SurvSHAP(t) and SurvLIME in a computational speed vs. accuracy trade-off. Finally, we apply these methods to medical data with multi-modal inputs, revealing relevant tabular features and visual patterns, as well as their temporal dynamics.

Cite this Paper

BibTeX
@InProceedings{pmlr-v267-langbein25a, title = {Gradient-based Explanations for Deep Learning Survival Models}, author = {Langbein, Sophie Hanna and Koenen, Niklas and Wright, Marvin N.}, booktitle = {Proceedings of the 42nd International Conference on Machine Learning}, pages = {32492--32522}, 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/langbein25a/langbein25a.pdf}, url = {https://proceedings.mlr.press/v267/langbein25a.html}, abstract = {Deep learning survival models often outperform classical methods in time-to-event predictions, particularly in personalized medicine, but their "black box" nature hinders broader adoption. We propose a framework for gradient-based explanation methods tailored to survival neural networks, extending their use beyond regression and classification. We analyze the implications of their theoretical assumptions for time-dependent explanations in the survival setting and propose effective visualizations incorporating the temporal dimension. Experiments on synthetic data show that gradient-based methods capture the magnitude and direction of local and global feature effects, including time dependencies. We introduce GradSHAP(t), a gradient-based counterpart to SurvSHAP(t), which outperforms SurvSHAP(t) and SurvLIME in a computational speed vs. accuracy trade-off. Finally, we apply these methods to medical data with multi-modal inputs, revealing relevant tabular features and visual patterns, as well as their temporal dynamics.} }
Endnote
%0 Conference Paper %T Gradient-based Explanations for Deep Learning Survival Models %A Sophie Hanna Langbein %A Niklas Koenen %A Marvin N. Wright %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-langbein25a %I PMLR %P 32492--32522 %U https://proceedings.mlr.press/v267/langbein25a.html %V 267 %X Deep learning survival models often outperform classical methods in time-to-event predictions, particularly in personalized medicine, but their "black box" nature hinders broader adoption. We propose a framework for gradient-based explanation methods tailored to survival neural networks, extending their use beyond regression and classification. We analyze the implications of their theoretical assumptions for time-dependent explanations in the survival setting and propose effective visualizations incorporating the temporal dimension. Experiments on synthetic data show that gradient-based methods capture the magnitude and direction of local and global feature effects, including time dependencies. We introduce GradSHAP(t), a gradient-based counterpart to SurvSHAP(t), which outperforms SurvSHAP(t) and SurvLIME in a computational speed vs. accuracy trade-off. Finally, we apply these methods to medical data with multi-modal inputs, revealing relevant tabular features and visual patterns, as well as their temporal dynamics.
APA
Langbein, S.H., Koenen, N. & Wright, M.N.. (2025). Gradient-based Explanations for Deep Learning Survival Models. Proceedings of the 42nd International Conference on Machine Learning, in Proceedings of Machine Learning Research 267:32492-32522 Available from https://proceedings.mlr.press/v267/langbein25a.html.

Related Material