Control Charts have been developed in 1920s to monitor mass production, but since then have found application in a number of fields, including epidemiology. Control charts may be statistical, economic or a hybrid of both. Unlike the common economic control charts which decide between doing something "now and never", dynamic ones decide between taking the action now or waiting for more information to make the optimal decision later. Deriving the optimal decision making strategies analytically is often impossible, thus application of a machine learning method called reinforcement learning to derive optimal control charts will be investigated within this project.
Supervisors
Supervisor: Elena Moltchanova
Key qualifications and skills
Mathematical statistics/Statistical inference, good programming skills (R preferred)
Does the project come with funding
No
Final date for receiving applications
Ongoing
Keywords
economic control charts; stochastic epidemic modeling; epidemiology; reinforcement learning