Predicting Stock Market Volatility by Bayesian Treed Gaussian Processes based on GARCH model

摘要：

we propose to predict financial volatility by a new treed Gaussian processes based on GARCH model. Three correlation. functions, isotropic exponential power, separable power and Matern families, are applied in the proposed hybrid treed GP models and stationary Gaussian processes. The empirical results show that the hybrid approaches generate better predictive capability than the stationary GARCH models; particularly, the treed Gaussian processes with Matern family correlation structure yields superior performance among the others.

关键词： bayesian tree gaussian process garch volatility

作者: PhichHang Ou Hengshan Wang

作者单位: Business School University of Shanghai for Science and Technology Shanghai, China

会议类型: 国际会议

会议名称: 2010 IEEE International Conference on Advanced Management Science(2010年IEEE高级管理科学国际会 IEEE ICAMS 2010)

会议地点: 成都

会议语种:英文

页码: 440-444

在线出版日期: 2010-07-09（万方平台首次上网日期，不代表论文的发表时间）