Proceedings of the Korean Information Science Society Conference (한국정보과학회:학술대회논문집)
- 2007.10c
- /
- Pages.485-489
- /
- 2007
- /
- 1598-5164(pISSN)
Outlier Robust Learning Algorithm for Gaussian Process Classification
가우시안 과정 분류를 위한 극단치에 강인한 학습 알고리즘
- Kim, Hyun-Chul (Yonsei University) ;
- Ghahramani, Zoubin (University of Cambridge)
- 김현철 (연세대학교) ;
- Published : 2007.10.26
Abstract
Gaussian process classifiers (GPCs) are fully statistical kernel classification models which have a latent function with Gaussian process prior Recently, EP approximation method has been proposed to infer the posterior over the latent function. It can have a special hyperparameter which can treat outliers potentially. In this paper, we propose the outlier robust algorithm which alternates EP and the hyperparameter updating until convergence. We also show its usefulness with the simulation results.
Keywords