Survival Filter: A Latent Timeseries Model for Joint Survival Analysis

Rajesh Ranganath Princeton University, Adler Perotte Columbia University, Noemie Elhadad Columbia University, David Blei Columbia University
Proceedings of the 31st Conference on Uncertainty in Artificial Intelligence, PMLR R13:742-751, 2015.

Abstract

Survival analysis is a core task in applied statistics, which models time-to-failure or time-to-event data. In the clinical domain, meaningful events can be the onset of different disease for a given patient. Because patients often have a wide range of diseases with complex interactions amongst them, it would be beneficial to model time to all diseases simultaneously. We propose and describe the survival filter model for this task, and apply it to a real-world, large dataset of longitudinal patient records. The model admits a scalable variational inference algorithm based on noisy gradients constructed from sampling the variational approximation. Experiments show that the survival filter model gives good predictive performance when compared to two baselines, and identifies clinically meaningful latent factors to represent diseases that co-occur in time.

Cite this Paper

BibTeX
@InProceedings{pmlr-vR13-university15p, title = {Survival Filter: A Latent Timeseries Model for Joint Survival Analysis}, author = {University, Rajesh Ranganath Princeton and University, Adler Perotte Columbia and University, Noemie Elhadad Columbia and University, David Blei Columbia}, booktitle = {Proceedings of the 31st Conference on Uncertainty in Artificial Intelligence}, pages = {742--751}, year = {2015}, editor = {Meila, Marina and Heskes, Tom}, volume = {R13}, series = {Proceedings of Machine Learning Research}, month = {12--16 Jul}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/r13/main/assets/university15p/university15p.pdf}, url = {https://proceedings.mlr.press/r13/university15p.html}, abstract = {Survival analysis is a core task in applied statistics, which models time-to-failure or time-to-event data. In the clinical domain, meaningful events can be the onset of different disease for a given patient. Because patients often have a wide range of diseases with complex interactions amongst them, it would be beneficial to model time to all diseases simultaneously. We propose and describe the survival filter model for this task, and apply it to a real-world, large dataset of longitudinal patient records. The model admits a scalable variational inference algorithm based on noisy gradients constructed from sampling the variational approximation. Experiments show that the survival filter model gives good predictive performance when compared to two baselines, and identifies clinically meaningful latent factors to represent diseases that co-occur in time.}, note = {Reissued by PMLR on 04 October 2026.} }
Endnote
%0 Conference Paper %T Survival Filter: A Latent Timeseries Model for Joint Survival Analysis %A Rajesh Ranganath Princeton University %A Adler Perotte Columbia University %A Noemie Elhadad Columbia University %A David Blei Columbia University %B Proceedings of the 31st Conference on Uncertainty in Artificial Intelligence %C Proceedings of Machine Learning Research %D 2015 %E Marina Meila %E Tom Heskes %F pmlr-vR13-university15p %I PMLR %P 742--751 %U https://proceedings.mlr.press/r13/university15p.html %V R13 %X Survival analysis is a core task in applied statistics, which models time-to-failure or time-to-event data. In the clinical domain, meaningful events can be the onset of different disease for a given patient. Because patients often have a wide range of diseases with complex interactions amongst them, it would be beneficial to model time to all diseases simultaneously. We propose and describe the survival filter model for this task, and apply it to a real-world, large dataset of longitudinal patient records. The model admits a scalable variational inference algorithm based on noisy gradients constructed from sampling the variational approximation. Experiments show that the survival filter model gives good predictive performance when compared to two baselines, and identifies clinically meaningful latent factors to represent diseases that co-occur in time. %Z Reissued by PMLR on 04 October 2026.
APA
University, R.R.P., University, A.P.C., University, N.E.C. & University, D.B.C.. (2015). Survival Filter: A Latent Timeseries Model for Joint Survival Analysis. Proceedings of the 31st Conference on Uncertainty in Artificial Intelligence, in Proceedings of Machine Learning Research R13:742-751 Available from https://proceedings.mlr.press/r13/university15p.html. Reissued by PMLR on 04 October 2026.

Related Material