Gaussian Process Models for Large-Scale Experiments

Gaussian process regression is a popular machine learning tool. But it is difficult to be applied to analyze large-scale experiment data with high dimension input (large \(p\)) and large sample size (large \(N\)). To overcome such issues, we propose a novel dimension reduction method that finds the optimal convex combination of low-dimension kernel functions for the GP model. It is shown that the proposed method is a significantly less computational and more accurate approximation of certain types of underlying functions. We also develop an active learning method based on the generalized Cook’s Distance is developed for the GP regression. It is more efficient than the standard random sampling method.

Event Topic

Computational Mathematics & Statistics