Machine Learning-Powered Mitigation Policy Optimization in Epidemiological Models

Jayaraman J. Thiagarajan, Rushil Anirudh, Peer-Timo Bremer, Timothy Germann, Sara Del Valle, Frederick Streitz
Proceedings of the 1st Workshop on Healthcare AI and COVID-19, ICML 2022, PMLR 184:63-72, 2022.

Abstract

A crucial aspect of managing a public health crisis is to effectively balance prevention and mitigation strategies, while taking their socio-economic impact into account. In particular, determining the influence of different non-pharmaceutical interventions (NPIs) on the effective use of public resources is an important problem, given the uncertainties on when a vaccine will be made available. In this paper, we propose a new approach for obtaining optimal policy recommendations based on Epicast models, which can characterize the disease progression under different interventions, and a look-ahead reward optimization strategy to choose the suitable NPI at different stages of an epidemic. Given the time delay inherent in any Epicast model and the exponential nature especially of an unmanaged epidemic, we find that such a look-ahead strategy infers non-trivial policies that adhere well to the constraints specified. Using two different Epicast models, namely SEIR and EpiCast, we evaluate the proposed algorithm to determine the optimal NPI policy, under a constraint on the number of daily new cases and the primary reward being the absence of restrictions.

Cite this Paper


BibTeX
@InProceedings{pmlr-v184-thiagarajan22a, title = {Machine Learning-Powered Mitigation Policy Optimization in Epidemiological Models}, author = {Thiagarajan, Jayaraman J. and Anirudh, Rushil and Bremer, Peer-Timo and Germann, Timothy and Del Valle, Sara and Streitz, Frederick}, booktitle = {Proceedings of the 1st Workshop on Healthcare AI and COVID-19, ICML 2022}, pages = {63--72}, year = {2022}, editor = {Xu, Peng and Zhu, Tingting and Zhu, Pengkai and Clifton, David A. and Belgrave, Danielle and Zhang, Yuanting}, volume = {184}, series = {Proceedings of Machine Learning Research}, month = {22 Jul}, publisher = {PMLR}, pdf = {https://proceedings.mlr.press/v184/thiagarajan22a/thiagarajan22a.pdf}, url = {https://proceedings.mlr.press/v184/thiagarajan22a.html}, abstract = {A crucial aspect of managing a public health crisis is to effectively balance prevention and mitigation strategies, while taking their socio-economic impact into account. In particular, determining the influence of different non-pharmaceutical interventions (NPIs) on the effective use of public resources is an important problem, given the uncertainties on when a vaccine will be made available. In this paper, we propose a new approach for obtaining optimal policy recommendations based on Epicast models, which can characterize the disease progression under different interventions, and a look-ahead reward optimization strategy to choose the suitable NPI at different stages of an epidemic. Given the time delay inherent in any Epicast model and the exponential nature especially of an unmanaged epidemic, we find that such a look-ahead strategy infers non-trivial policies that adhere well to the constraints specified. Using two different Epicast models, namely SEIR and EpiCast, we evaluate the proposed algorithm to determine the optimal NPI policy, under a constraint on the number of daily new cases and the primary reward being the absence of restrictions.} }
Endnote
%0 Conference Paper %T Machine Learning-Powered Mitigation Policy Optimization in Epidemiological Models %A Jayaraman J. Thiagarajan %A Rushil Anirudh %A Peer-Timo Bremer %A Timothy Germann %A Sara Del Valle %A Frederick Streitz %B Proceedings of the 1st Workshop on Healthcare AI and COVID-19, ICML 2022 %C Proceedings of Machine Learning Research %D 2022 %E Peng Xu %E Tingting Zhu %E Pengkai Zhu %E David A. Clifton %E Danielle Belgrave %E Yuanting Zhang %F pmlr-v184-thiagarajan22a %I PMLR %P 63--72 %U https://proceedings.mlr.press/v184/thiagarajan22a.html %V 184 %X A crucial aspect of managing a public health crisis is to effectively balance prevention and mitigation strategies, while taking their socio-economic impact into account. In particular, determining the influence of different non-pharmaceutical interventions (NPIs) on the effective use of public resources is an important problem, given the uncertainties on when a vaccine will be made available. In this paper, we propose a new approach for obtaining optimal policy recommendations based on Epicast models, which can characterize the disease progression under different interventions, and a look-ahead reward optimization strategy to choose the suitable NPI at different stages of an epidemic. Given the time delay inherent in any Epicast model and the exponential nature especially of an unmanaged epidemic, we find that such a look-ahead strategy infers non-trivial policies that adhere well to the constraints specified. Using two different Epicast models, namely SEIR and EpiCast, we evaluate the proposed algorithm to determine the optimal NPI policy, under a constraint on the number of daily new cases and the primary reward being the absence of restrictions.
APA
Thiagarajan, J.J., Anirudh, R., Bremer, P., Germann, T., Del Valle, S. & Streitz, F.. (2022). Machine Learning-Powered Mitigation Policy Optimization in Epidemiological Models. Proceedings of the 1st Workshop on Healthcare AI and COVID-19, ICML 2022, in Proceedings of Machine Learning Research 184:63-72 Available from https://proceedings.mlr.press/v184/thiagarajan22a.html.

Related Material