南阳师范学院学报2016，Vol.15Issue(12)：12-17,6.
信噪比对非参数回归模型拟合效果的数值模拟
Numerical simulation for the fitting effect of signal-to-noise ratio on nonparametric regressive models
摘要
Abstract
To illustrate the fitting effect of signal-to-noise ratio (SNR) on nonparametric regression models, a Monte Carlo numerical simulation study is conducted under the case of fixed design.In the simulation, cubic spline with the B-spline basis to construct nonparametric estimates is used, and the optimal number of interior knots is selected by AIC or BIC.The results show that the larger the SNR, the smaller the mean and standard deviation of the Square-Root of Average Squared Error for the estimates of mean function, and the smaller the Mean Squared Error and Mean Absolute Error of response variable fitting.关键词
非参数回归模型/多项式样条方法/信噪比/MonteCarlo数值模拟Key words
nonparametric regressive models/polynomial spline estimation/signal-to-noise ratio/Monte Carlo numerical simulation分类
数理科学引用本文复制引用
程芳,武新乾,杨万才..信噪比对非参数回归模型拟合效果的数值模拟[J].南阳师范学院学报,2016,15(12):12-17,6.基金项目
国家自然科学基金(11501167) （11501167）
河南省国际科技合作计划项目(134300510034) （134300510034）
河南省软科学研究项目(152400410152) （152400410152）
