Adaptive Discretization for Event PredicTion (ADEPT)

Jimmy Hickey, Ricardo Henao, Daniel Wojdyla, Michael Pencina, Matthew Engelhard
Proceedings of The 27th International Conference on Artificial Intelligence and Statistics, PMLR 238:1351-1359, 2024.

Abstract

Recently developed survival analysis methods improve upon existing approaches by predicting the probability of event occurrence in each of a number pre-specified (discrete) time intervals. By avoiding placing strong parametric assumptions on the event density, this approach tends to improve prediction performance, particularly when data are plentiful. However, in clinical settings with limited available data, it is often preferable to judiciously partition the event time space into a limited number of intervals well suited to the prediction task at hand. In this work, we develop Adaptive Discretization for Event PredicTion (ADEPT) to learn from data a set of cut points defining such a partition. We show that in two simulated datasets, we are able to recover intervals that match the underlying generative model. We then demonstrate improved prediction performance on three real-world observational datasets, including a large, newly harmonized stroke risk prediction dataset. Finally, we argue that our approach facilitates clinical decision-making by suggesting time intervals that are most appropriate for each task, in the sense that they facilitate more accurate risk prediction.

Cite this Paper

BibTeX
@InProceedings{pmlr-v238-hickey24a, title = { Adaptive Discretization for Event PredicTion {(ADEPT)} }, author = {Hickey, Jimmy and Henao, Ricardo and Wojdyla, Daniel and Pencina, Michael and Engelhard, Matthew}, booktitle = {Proceedings of The 27th International Conference on Artificial Intelligence and Statistics}, pages = {1351--1359}, 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/hickey24a/hickey24a.pdf}, url = {https://proceedings.mlr.press/v238/hickey24a.html}, abstract = { Recently developed survival analysis methods improve upon existing approaches by predicting the probability of event occurrence in each of a number pre-specified (discrete) time intervals. By avoiding placing strong parametric assumptions on the event density, this approach tends to improve prediction performance, particularly when data are plentiful. However, in clinical settings with limited available data, it is often preferable to judiciously partition the event time space into a limited number of intervals well suited to the prediction task at hand. In this work, we develop Adaptive Discretization for Event PredicTion (ADEPT) to learn from data a set of cut points defining such a partition. We show that in two simulated datasets, we are able to recover intervals that match the underlying generative model. We then demonstrate improved prediction performance on three real-world observational datasets, including a large, newly harmonized stroke risk prediction dataset. Finally, we argue that our approach facilitates clinical decision-making by suggesting time intervals that are most appropriate for each task, in the sense that they facilitate more accurate risk prediction. } }
Endnote
%0 Conference Paper %T Adaptive Discretization for Event PredicTion (ADEPT) %A Jimmy Hickey %A Ricardo Henao %A Daniel Wojdyla %A Michael Pencina %A Matthew Engelhard %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-hickey24a %I PMLR %P 1351--1359 %U https://proceedings.mlr.press/v238/hickey24a.html %V 238 %X Recently developed survival analysis methods improve upon existing approaches by predicting the probability of event occurrence in each of a number pre-specified (discrete) time intervals. By avoiding placing strong parametric assumptions on the event density, this approach tends to improve prediction performance, particularly when data are plentiful. However, in clinical settings with limited available data, it is often preferable to judiciously partition the event time space into a limited number of intervals well suited to the prediction task at hand. In this work, we develop Adaptive Discretization for Event PredicTion (ADEPT) to learn from data a set of cut points defining such a partition. We show that in two simulated datasets, we are able to recover intervals that match the underlying generative model. We then demonstrate improved prediction performance on three real-world observational datasets, including a large, newly harmonized stroke risk prediction dataset. Finally, we argue that our approach facilitates clinical decision-making by suggesting time intervals that are most appropriate for each task, in the sense that they facilitate more accurate risk prediction.
APA
Hickey, J., Henao, R., Wojdyla, D., Pencina, M. & Engelhard, M.. (2024). Adaptive Discretization for Event PredicTion (ADEPT) . Proceedings of The 27th International Conference on Artificial Intelligence and Statistics, in Proceedings of Machine Learning Research 238:1351-1359 Available from https://proceedings.mlr.press/v238/hickey24a.html.

Related Material