Bayesian Functional Optimisation with Shape Prior


  • Pratibha Vellanki Deakin University
  • Santu Rana Deakin University
  • Sunil Gupta Deakin University
  • David Rubin de Celis Leal Deakin University
  • Alessandra Sutti Deakin University
  • Murray Height Deakin University
  • Svetha Venkatesh Deakin University



Real world experiments are expensive, and thus it is important to reach a target in a minimum number of experiments. Experimental processes often involve control variables that change over time. Such problems can be formulated as functional optimisation problem. We develop a novel Bayesian optimisation framework for such functional optimisation of expensive black-box processes. We represent the control function using Bernstein polynomial basis and optimise in the coefficient space. We derive the theory and practice required to dynamically adjust the order of the polynomial degree, and show how prior information about shape can be integrated. We demonstrate the effectiveness of our approach for short polymer fibre design and optimising learning rate schedules for deep networks.




How to Cite

Vellanki, P., Rana, S., Gupta, S., Rubin de Celis Leal, D., Sutti, A., Height, M., & Venkatesh, S. (2019). Bayesian Functional Optimisation with Shape Prior. Proceedings of the AAAI Conference on Artificial Intelligence, 33(01), 1617-1624.



AAAI Technical Track: Constraint Satisfaction and Optimization