Multi-Task Time Series Analysis applied to Drug Response Modelling

Alexander Bird; Christopher K I Williams; Christopher  Hawthorne

Multi-Task Time Series Analysis applied to Drug Response Modelling

Alexander Bird, Christopher K I Williams, Christopher Hawthorne

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

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.

Original language	English
Title of host publication	Proceedings of the Twenty-Second International Conference on Artificial Intelligence and Statistics
Editors	Neil Lawrence, Mark Reid
Publisher	PMLR
Number of pages	10
Volume	89
Publication status	E-pub ahead of print - 18 Apr 2019
Event	22nd International Conference on Artificial Intelligence and Statistics - Naha, Japan Duration: 16 Apr 2019 → 18 Apr 2019 https://www.aistats.org/

Publication series

Name	Proceedings of Machine Learning Research
Publisher	PMLR
Volume	89
ISSN (Electronic)	2640-3498

Conference

Conference	22nd International Conference on Artificial Intelligence and Statistics
Abbreviated title	AISTATS 2019
Country/Territory	Japan
City	Naha
Period	16/04/19 → 18/04/19
Internet address	https://www.aistats.org/

Access to Document

Multi-Task Time Series_BIRD_DoA-241218_AFVAccepted author manuscript, 768 KB

http://proceedings.mlr.press/v89/bird19a.html

ATI Faculty Fellowship
Williams, C. (Principal Investigator)
UK-based charities
1/01/16 → 31/07/20
Project: Research
Improving decision support for trating arterial hypotension in adult patients during their management in intensive care
Williams, C. (Principal Investigator)
UK central government bodies/local authorities, health and hospital authorities
1/05/13 → 30/04/15
Project: Research

Cite this

Bird, A., Williams, C. K. I., & Hawthorne, C. (2019). Multi-Task Time Series Analysis applied to Drug Response Modelling. In N. Lawrence, & M. Reid (Eds.), Proceedings of the Twenty-Second International Conference on Artificial Intelligence and Statistics (Vol. 89). (Proceedings of Machine Learning Research; Vol. 89). PMLR. Advance online publication. http://proceedings.mlr.press/v89/bird19a.html

@inproceedings{3d9a441b259343e4ae773e09d602ee20,

title = "Multi-Task Time Series Analysis applied to Drug Response Modelling",

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.",

author = "Alexander Bird and Williams, {Christopher K I} and Christopher Hawthorne",

year = "2019",

month = apr,

day = "18",

language = "English",

volume = "89",

series = "Proceedings of Machine Learning Research",

publisher = "PMLR",

editor = "Neil Lawrence and Mark Reid",

booktitle = "Proceedings of the Twenty-Second International Conference on Artificial Intelligence and Statistics",

note = "22nd International Conference on Artificial Intelligence and Statistics, AISTATS 2019 ; Conference date: 16-04-2019 Through 18-04-2019",

url = "https://www.aistats.org/",

}

Bird, A, Williams, CKI & Hawthorne, C 2019, Multi-Task Time Series Analysis applied to Drug Response Modelling. in N Lawrence & M Reid (eds), Proceedings of the Twenty-Second International Conference on Artificial Intelligence and Statistics. vol. 89, Proceedings of Machine Learning Research, vol. 89, PMLR, 22nd International Conference on Artificial Intelligence and Statistics, Naha, Japan, 16/04/19. <http://proceedings.mlr.press/v89/bird19a.html>

Multi-Task Time Series Analysis applied to Drug Response Modelling. / Bird, Alexander; Williams, Christopher K I; Hawthorne, Christopher .
Proceedings of the Twenty-Second International Conference on Artificial Intelligence and Statistics. ed. / Neil Lawrence; Mark Reid. Vol. 89 PMLR, 2019. (Proceedings of Machine Learning Research; Vol. 89).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

TY - GEN

T1 - Multi-Task Time Series Analysis applied to Drug Response Modelling

AU - Bird, Alexander

AU - Williams, Christopher K I

AU - Hawthorne, Christopher

PY - 2019/4/18

Y1 - 2019/4/18

N2 - 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.

AB - 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.

M3 - Conference contribution

VL - 89

T3 - Proceedings of Machine Learning Research

BT - Proceedings of the Twenty-Second International Conference on Artificial Intelligence and Statistics

A2 - Lawrence, Neil

A2 - Reid, Mark

PB - PMLR

T2 - 22nd International Conference on Artificial Intelligence and Statistics

Y2 - 16 April 2019 through 18 April 2019

ER -

Multi-Task Time Series Analysis applied to Drug Response Modelling

Abstract

Publication series

Conference

Access to Document

Fingerprint

Projects

ATI Faculty Fellowship

Improving decision support for trating arterial hypotension in adult patients during their management in intensive care

Cite this