论文信息 - Sparse Representation for Gaussian Process Models

Sparse Representation for Gaussian Process Models

We develop an approach for a sparse representation for Gaussian Process (GP) models in order to overcome the limitations of GPs caused by large data sets. The method is based on a combination of a Bayesian online algorithm together with a sequential construction of a relevant subsample of the data which fully specifies the prediction of the model. Experimental results on toy examples and large real-world data sets indicate the efficiency of the approach.

Manfred Opper | Lehel Csató | M. Opper | L. Csató

[1] J. Davenport. Editor , 1960 .

[2] Carl E. Rasmussen,et al. In Advances in Neural Information Processing Systems , 2011 .

[3] M. Gibbs,et al. Efficient implementation of gaussian processes , 1997 .

[4] Manfred Opper,et al. Finite-Dimensional Approximation of Gaussian Processes , 1998, NIPS.

[5] David Saad,et al. On-Line Learning in Neural Networks , 1999 .

[6] Matthias W. Seeger,et al. Bayesian Model Selection for Support Vector Machines, Gaussian Processes and Other Kernel Classifiers , 1999, NIPS.

[7] B. Schölkopf,et al. Advances in kernel methods: support vector learning , 1999 .

[8] Gunnar Rätsch,et al. Input space versus feature space in kernel-based methods , 1999, IEEE Trans. Neural Networks.

[9] Michael E. Tipping. The Relevance Vector Machine , 1999, NIPS.

[10] Ole Winther,et al. Efficient Approaches to Gaussian Process Classification , 1999, NIPS.

[11] Volker Tresp,et al. A Bayesian Committee Machine , 2000, Neural Computation.

[12] Vladimir N. Vapnik,et al. The Nature of Statistical Learning Theory , 2000, Statistics for Engineering and Information Science.

[13] Lemm,et al. Bayesian approach to inverse quantum statistics , 2000, Physical review letters.