Speaker:Dr Robert Gramacy (Stats Lab)
Venue: Old Combination Room, Trinity College
Time: 03/03/2008 20:30, drinks from 20:15

The talk is available.

Computer experiments often require dense sweeps over input parameters to obtain a qualitative understanding of their response. However, such sweeps are unnecessary in regions where the response is easily predicted; well-chosen designs could allow a mapping of the response with far fewer simulation runs. I explore a modern approach that couples two standard regression models: Gaussian processes and treed partitioning. A Bayesian perspective yields an explicit measure of (nonstationary) predictive uncertainty that can be used to guide sampling. The methods will be illustrated through several examples, including a motivating example which involves the computational fluid dynamics of a NASA re-entry vehicle.