摘要
针对影响台风最大风速的输入变量较多以及输入变量与输出变量之间的非线性变化特点,首先计算各个输入变量与输出变量间的互信息,这些互信息间接地反映了各个输入变量与输出变量间的相关性;然后根据t检验法确定一个阈值,对于互信息小于阈值的输入变量作不相关变量处理,筛选出最佳的模型输入变量;最后采用高斯过程回归模型对筛选后的样本集进行拟合,在贝叶斯非参数建模的框架下,确定高斯过程回归模型的协方差函数。仿真结果表明,所得高斯过程模型能够满足绝对误差的预定要求,且具有较大的实用价值。
In the prediction model for the maximum wind speed of typhoon, the number of the input variables is very large and the output variable is a nonlinear function of the input variables. Firstly, this paper calculated the mutual information between each input variable and the output variable, because the mutual information indirectly reflected the correlation between the in- put and output variables. Determining a threshold on the mutual information with a t-test,it discarded the input variables whose mutual information with the output variable is less than the threshold. After it selected the optimal model input variables, a Gaussian process regression model was fitted to the selected sample set. At the same time it determined the hyper-parameters in the covariance function under the Bayesian nonparametric framework. The simulation result shows that the Gaussian process re- gression model has met the predetermined requirements on absolute error, and it has great practical values.
出处
《计算机应用研究》
CSCD
北大核心
2015年第1期59-62,共4页
Application Research of Computers
基金
国家自然科学基金资助项目(61273070)
江苏省自然科学基金资助项目(BK2011003)
高等学校学科创新引智计划资助项目(B12018)
江苏高校优势学科建设工程资助项目
江南大学自主科研计划青年基金资助项目(JUSRP11132)
关键词
高斯过程
回归分析
贝叶斯非参数模型
特征选择
互信息
Gaussian process
regression analysis
Bayesian nonparametric model
feature selection
mutual information
