The Bayesian a posterior probability is a very important element in pattern recognition. In classification problems, the posterior probabilities reflect the uncertainty of assessing an example to particular class. Such residual information will be useful for more deep understanding or analysis of examples. In this paper, we propose a nonlinear discriminant analysis based on the probabilistic estimation of the Gaussian mixture model (GMM). We use GMM to estimate the Bayesian a posterior probabilities of any classification problems. Then we use posterior probabilities estimated by GMM to construct discriminative kernel function. The performance of the proposed kernel function is confirmed by several experiments using UCI machine learning repository. © 2014 Springer-Verlag Berlin Heidelberg.
CITATION STYLE
Hidaka, A., & Kurita, T. (2014). Nonlinear discriminant analysis based on probability estimation by gaussian mixture model. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 8621 LNCS, pp. 133–142). Springer Verlag. https://doi.org/10.1007/978-3-662-44415-3_14
Mendeley helps you to discover research relevant for your work.