Component-Wise Composite Likelihood Distillation for Censored Time-to-Event Data

Feiyang Deng, Lingfeng Luo, Jiayu Zhou, Kevin He
Proceedings of the 43rd International Conference on Machine Learning, PMLR 306:24170-24190, 2026.

Abstract

Accurate survival modeling in biomedical studies is often hindered by rare events, limited effective sample sizes, and settings with limited or partially observed information (e.g., covariates of interest that are difficult or expensive to collect, highly structured sampling designs, or nuisance parameters omitted by conditioning). Knowledge distillation can leverage external predictive information without sharing individual-level data, but existing approaches are largely built for fully specified likelihoods or probability-based survival models and do not extend to settings where outcome distributions are only partially specified. To address this challenge, we propose a knowledge distillation framework based on a composite-likelihood Kullback–Leibler divergence that aligns teacher and student models within components. Our key insight is that, although composite likelihoods do not define a global outcome distribution, each likelihood component induces a well-defined probability model on its restricted outcome space, enabling a principled KL divergence. Simulation studies and biomedical case studies show improved discrimination and predictive accuracy in rare-event, heterogeneous settings without requiring access to external individual-level data.

Cite this Paper

BibTeX
@InProceedings{pmlr-v306-deng26p, title = {Component-Wise Composite Likelihood Distillation for Censored Time-to-Event Data}, author = {Deng, Feiyang and Luo, Lingfeng and Zhou, Jiayu and He, Kevin}, booktitle = {Proceedings of the 43rd International Conference on Machine Learning}, pages = {24170--24190}, year = {2026}, editor = {Zhang, Tong and Dudik, Miroslav and Jaggi, Martin and Agarwal, Alekh and Li, Sharon and Schuurmans, Dale and Zhu, Jerry and Berkenkamp, Felix and Dong, Hanze and Bietti, Alberto}, volume = {306}, series = {Proceedings of Machine Learning Research}, month = {06--11 Jul}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v306/main/assets/deng26p/deng26p.pdf}, url = {https://proceedings.mlr.press/v306/deng26p.html}, abstract = {Accurate survival modeling in biomedical studies is often hindered by rare events, limited effective sample sizes, and settings with limited or partially observed information (e.g., covariates of interest that are difficult or expensive to collect, highly structured sampling designs, or nuisance parameters omitted by conditioning). Knowledge distillation can leverage external predictive information without sharing individual-level data, but existing approaches are largely built for fully specified likelihoods or probability-based survival models and do not extend to settings where outcome distributions are only partially specified. To address this challenge, we propose a knowledge distillation framework based on a composite-likelihood Kullback–Leibler divergence that aligns teacher and student models within components. Our key insight is that, although composite likelihoods do not define a global outcome distribution, each likelihood component induces a well-defined probability model on its restricted outcome space, enabling a principled KL divergence. Simulation studies and biomedical case studies show improved discrimination and predictive accuracy in rare-event, heterogeneous settings without requiring access to external individual-level data.} }
Endnote
%0 Conference Paper %T Component-Wise Composite Likelihood Distillation for Censored Time-to-Event Data %A Feiyang Deng %A Lingfeng Luo %A Jiayu Zhou %A Kevin He %B Proceedings of the 43rd International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2026 %E Tong Zhang %E Miroslav Dudik %E Martin Jaggi %E Alekh Agarwal %E Sharon Li %E Dale Schuurmans %E Jerry Zhu %E Felix Berkenkamp %E Hanze Dong %E Alberto Bietti %F pmlr-v306-deng26p %I PMLR %P 24170--24190 %U https://proceedings.mlr.press/v306/deng26p.html %V 306 %X Accurate survival modeling in biomedical studies is often hindered by rare events, limited effective sample sizes, and settings with limited or partially observed information (e.g., covariates of interest that are difficult or expensive to collect, highly structured sampling designs, or nuisance parameters omitted by conditioning). Knowledge distillation can leverage external predictive information without sharing individual-level data, but existing approaches are largely built for fully specified likelihoods or probability-based survival models and do not extend to settings where outcome distributions are only partially specified. To address this challenge, we propose a knowledge distillation framework based on a composite-likelihood Kullback–Leibler divergence that aligns teacher and student models within components. Our key insight is that, although composite likelihoods do not define a global outcome distribution, each likelihood component induces a well-defined probability model on its restricted outcome space, enabling a principled KL divergence. Simulation studies and biomedical case studies show improved discrimination and predictive accuracy in rare-event, heterogeneous settings without requiring access to external individual-level data.
APA
Deng, F., Luo, L., Zhou, J. & He, K.. (2026). Component-Wise Composite Likelihood Distillation for Censored Time-to-Event Data. Proceedings of the 43rd International Conference on Machine Learning, in Proceedings of Machine Learning Research 306:24170-24190 Available from https://proceedings.mlr.press/v306/deng26p.html.

Related Material