Cookies on this website

We use cookies to ensure that we give you the best experience on our website. If you click 'Accept all cookies' we'll assume that you are happy to receive all cookies and you won't see this message again. If you click 'Reject all non-essential cookies' only necessary cookies providing core functionality such as security, network management, and accessibility will be enabled. Click 'Find out more' for information on how to change your cookie settings.


In-patient hospital data presents unique challenges for time series analysis, including the sparsity and irregularity of observations for each patient and the heterogeneous patient responses to interventions. In this talk, I will present a multi-output Gaussian process regression model for patient time series data that captures the state of a patient and uncertainty in this state across four vital signs and 20 lab tests in a patient-specific way. We build on top of this model a reinforcement learning approach to assist doctors to wean patients from a mechanical ventilator. Finally, I show how prior work with time series associations may be used with these data to identify patients with genetically-mediated responses to specific interventions. I will conclude with directions for future work.