Pseudo-density Estimation for Clustering with Gaussian Processes
Gaussian processes (GP) provide a kernel machine framework. They have been mainly applied to regression and classification. We propose a pseudo-density estimation method based on the information of variance functions of GPs, which relates to the density of the data points. We also show how the constructed pseudo-density can be applied to clustering. Through simulation we show that the topology of the pseudo-density represents the clustering information well with promising results.
KeywordsGaussian Process Sample Function Gaussian Process Regression Gaussian Process Model Kernel Machine
Unable to display preview. Download preview PDF.
- 2.Bezdek, J.C.: Pattern Recognition with Fuzzy Objective Function Algoritms. Plenum Press, New York (1981)Google Scholar
- 5.Dunn, J.C.: A fuzzy relative of the isodata process and its use in detecting compact well-separated clusters. Journal of Cybernetics, 32–57 (1973)Google Scholar
- 12.Neal, R.M.: Regression and Classification Using Gaussian Process Priors. Bayesian Statistics 6, 465–501 (1998)Google Scholar
- 13.Rasmussen, C.E.: Evaluation of Gaussian Processes and Other Methods for Non-Linear Regression. PhD Thesis University of Toronto (1996)Google Scholar
- 14.Williams, C.K.I., Rasmussen, C.E.: Gaussian Processes for Regression. NIPS 8, 514–520 (1995)Google Scholar