Hierarchical Gaussian Process Regression

[edit]

Sunho Park, Seungjin Choi ;
Proceedings of 2nd Asian Conference on Machine Learning, PMLR 13:95-110, 2010.

Abstract

We address an approximation method for Gaussian process (GP) regression, where we approximate covariance by a block matrix such that diagonal blocks are calculated exactly while off-diagonal blocks are approximated. Partitioning input data points, we present a two-layer hierarchical model for GP regression, where prototypes of clusters in the upper layer are involved for coarse modeling by a GP and data points in each cluster in the lower layer are involved for fine modeling by an individual GP whose prior mean is given by the corresponding prototype and covariance is parameterized by data points in the partition. In this hierarchical model, integrating out latent variables in the upper layer leads to a block covariance matrix, where diagonal blocks contain similarities between data points in the same partition and off-diagonal blocks consist of approximate similarities calculated using prototypes. This particular structure of the covariance matrix divides the full GP into a pieces of manageable sub-problems whose complexity scales with the number of data points in a partition. In addition, our hierarchical GP regression (HGPR) is also useful for cases where partitions of data reveal different characteristics. Experiments on several benchmark datasets confirm the useful behavior of our method.

Related Material