Multi-Task Time Series Analysis applied to Drug Response Modelling

Alex Bird, Chris Williams, Christopher Hawthorne
Proceedings of the Twenty-Second International Conference on Artificial Intelligence and Statistics, PMLR 89:2174-2183, 2019.

Abstract

Time series models such as dynamical systems are frequently fitted to a cohort of data, ignoring variation between individual entities such as patients. In this paper we show how these models can be personalised to an individual level while retaining statistical power, via use of multi-task learning (MTL). To our knowledge this is a novel development of MTL which applies to time series both with and without control inputs. The modelling framework is demonstrated on a physiological drug response problem which results in improved predictive accuracy and uncertainty estimation over existing state-of-the-art models.

Cite this Paper

BibTeX
@InProceedings{pmlr-v89-bird19a, title = {Multi-Task Time Series Analysis applied to Drug Response Modelling}, author = {Bird, Alex and Williams, Chris and Hawthorne, Christopher}, booktitle = {Proceedings of the Twenty-Second International Conference on Artificial Intelligence and Statistics}, pages = {2174--2183}, year = {2019}, editor = {Chaudhuri, Kamalika and Sugiyama, Masashi}, volume = {89}, series = {Proceedings of Machine Learning Research}, month = {16--18 Apr}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v89/bird19a/bird19a.pdf}, url = {https://proceedings.mlr.press/v89/bird19a.html}, abstract = {Time series models such as dynamical systems are frequently fitted to a cohort of data, ignoring variation between individual entities such as patients. In this paper we show how these models can be personalised to an individual level while retaining statistical power, via use of multi-task learning (MTL). To our knowledge this is a novel development of MTL which applies to time series both with and without control inputs. The modelling framework is demonstrated on a physiological drug response problem which results in improved predictive accuracy and uncertainty estimation over existing state-of-the-art models.} }
Endnote
%0 Conference Paper %T Multi-Task Time Series Analysis applied to Drug Response Modelling %A Alex Bird %A Chris Williams %A Christopher Hawthorne %B Proceedings of the Twenty-Second International Conference on Artificial Intelligence and Statistics %C Proceedings of Machine Learning Research %D 2019 %E Kamalika Chaudhuri %E Masashi Sugiyama %F pmlr-v89-bird19a %I PMLR %P 2174--2183 %U https://proceedings.mlr.press/v89/bird19a.html %V 89 %X Time series models such as dynamical systems are frequently fitted to a cohort of data, ignoring variation between individual entities such as patients. In this paper we show how these models can be personalised to an individual level while retaining statistical power, via use of multi-task learning (MTL). To our knowledge this is a novel development of MTL which applies to time series both with and without control inputs. The modelling framework is demonstrated on a physiological drug response problem which results in improved predictive accuracy and uncertainty estimation over existing state-of-the-art models.
APA
Bird, A., Williams, C. & Hawthorne, C.. (2019). Multi-Task Time Series Analysis applied to Drug Response Modelling. Proceedings of the Twenty-Second International Conference on Artificial Intelligence and Statistics, in Proceedings of Machine Learning Research 89:2174-2183 Available from https://proceedings.mlr.press/v89/bird19a.html.

Related Material