SUPPORT VECTOR REGRESSION VIA MCMC WITHIN EVIDENCE FRAMEWORK  

作　　者：Zhou Yatong Li Jin Sun Jiancheng Zhang Bolun 

机构地区：[1]School of Information Engineering, Hebei University of Technology [2]School of Software and Communication Engineering, Jiangxi University of Finance and Economics

出　　处：《Journal of Electronics(China)》2012年第6期530-533,共4页电子科学学刊（英文版）

基　　金：Supported by the National Natural Science Foundation of China (No. 60972106, 61072103);China Postdoctoral Science Foundation (No. 20090450750)

摘　　要：This paper proposes a novel approach, Markov Chain Monte Carlo (MCMC) sampling approximation, to deal with intractable high-dimension integral in the evidence framework applied to Support Vector Regression (SVR). Unlike traditional variational or mean field method, the proposed approach follows the idea of MCMC, firstly draws some samples from the posterior distribution on SVR's weight vector, and then approximates the expected output integrals by finite sums. Experimental results show the proposed approach is feasible and robust to noise. It also shows the performance of proposed approach and Relevance Vector Machine (RVM) is comparable under the noise circumstances. They give better robustness compared to standard SVR.This paper proposes a novel approach, Markov Chain Monte Carlo （MCMC） sampling approximation, to deal with intractable high-dimension integral in the evidence framework applied to Support Vector Regression （SVR）. Unlike traditional variational or mean field method, the proposed approach follows the idea of MCMC, firstly draws some samples from the posterior distribution on SVR＇s weight vector, and then approximates the expected output integrals by finite sums. Experi- mental results show the proposed approach is feasible and robust to noise. It also shows the per- formance of proposed approach and Relevance Vector Machine （RVM） is comparable under the noise circumstances. They give better robustness compared to standard SVR.

关 键 词：Support Vector Regression (SVR) Markov Chain Monte Carlo (MCMC) Evidence Framework (EF) Noise 

分 类 号：TP751[自动化与计算机技术—检测技术与自动化装置]