Harmonic-Mean Cox Models: A Ruler for Equal Attention to Risk

Xuejian Wang, Wenbin Zhang, Aishwarya Jadhav, Jeremy Weiss
Proceedings of AAAI Spring Symposium on Survival Prediction - Algorithms, Challenges, and Applications 2021, PMLR 146:171-183, 2021.

Abstract

Survival analysis models are necessary for clinical forecasting with data censorship. Implicitly, existing works focus on the individuals with higher risks while lower risk individuals are poorly characterized. Developing survival models to represent different risk individuals equally is a challenging task but of great importance for providing accurate risk assessments across levels of risk. Here, we characterize this problem and propose an adjusted log-likelihood formulation as the new objective for survival prognostication. Several models are then proposed based on the newly designed optimization objective function which produce risks that count individuals “equally” on risk ratios thus providing representative attention to individuals of varying risk. Extensive experiments on multiple real-world datasets demonstrate the benefits of the proposed approach.

Cite this Paper

BibTeX
@InProceedings{pmlr-v146-wang21a, title = {Harmonic-Mean Cox Models: A Ruler for Equal Attention to Risk}, author = {Wang, Xuejian and Zhang, Wenbin and Jadhav, Aishwarya and Weiss, Jeremy}, booktitle = {Proceedings of AAAI Spring Symposium on Survival Prediction - Algorithms, Challenges, and Applications 2021}, pages = {171--183}, year = {2021}, editor = {Greiner, Russell and Kumar, Neeraj and Gerds, Thomas Alexander and van der Schaar, Mihaela}, volume = {146}, series = {Proceedings of Machine Learning Research}, month = {22--24 Mar}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v146/wang21a/wang21a.pdf}, url = {https://proceedings.mlr.press/v146/wang21a.html}, abstract = {Survival analysis models are necessary for clinical forecasting with data censorship. Implicitly, existing works focus on the individuals with higher risks while lower risk individuals are poorly characterized. Developing survival models to represent different risk individuals equally is a challenging task but of great importance for providing accurate risk assessments across levels of risk. Here, we characterize this problem and propose an adjusted log-likelihood formulation as the new objective for survival prognostication. Several models are then proposed based on the newly designed optimization objective function which produce risks that count individuals “equally” on risk ratios thus providing representative attention to individuals of varying risk. Extensive experiments on multiple real-world datasets demonstrate the benefits of the proposed approach.} }
Endnote
%0 Conference Paper %T Harmonic-Mean Cox Models: A Ruler for Equal Attention to Risk %A Xuejian Wang %A Wenbin Zhang %A Aishwarya Jadhav %A Jeremy Weiss %B Proceedings of AAAI Spring Symposium on Survival Prediction - Algorithms, Challenges, and Applications 2021 %C Proceedings of Machine Learning Research %D 2021 %E Russell Greiner %E Neeraj Kumar %E Thomas Alexander Gerds %E Mihaela van der Schaar %F pmlr-v146-wang21a %I PMLR %P 171--183 %U https://proceedings.mlr.press/v146/wang21a.html %V 146 %X Survival analysis models are necessary for clinical forecasting with data censorship. Implicitly, existing works focus on the individuals with higher risks while lower risk individuals are poorly characterized. Developing survival models to represent different risk individuals equally is a challenging task but of great importance for providing accurate risk assessments across levels of risk. Here, we characterize this problem and propose an adjusted log-likelihood formulation as the new objective for survival prognostication. Several models are then proposed based on the newly designed optimization objective function which produce risks that count individuals “equally” on risk ratios thus providing representative attention to individuals of varying risk. Extensive experiments on multiple real-world datasets demonstrate the benefits of the proposed approach.
APA
Wang, X., Zhang, W., Jadhav, A. & Weiss, J.. (2021). Harmonic-Mean Cox Models: A Ruler for Equal Attention to Risk. Proceedings of AAAI Spring Symposium on Survival Prediction - Algorithms, Challenges, and Applications 2021, in Proceedings of Machine Learning Research 146:171-183 Available from https://proceedings.mlr.press/v146/wang21a.html.

Related Material