采用次序逻辑回归模型 ordianl logistic regression model
目的:比较用于处理包含未检出值的痕量测定数据的次序统计量回归方法与经典替换方法的估计效果。
Objective: the purpose of this study was to compare the performance of regression on order statistics(ROS) and substitution methods in estimation of nondetects.
结果:在数据服从正态和对数正态分布时,次序统计量回归效果明显优于简单替换法;资料服从非正态分布资料时,次序统计量回归方法没有明显优势。
Objective: the purpose of this study was to compare the performance of regression on order statistics(ROS) and substitution methods in estimation of nondetects.
应用线性回归分析和移动平均理论,对按时间次序排列的单一数据序列,给出了一种线性移动自回归预测模型,并对原始数据受不确定因素影响而产生的随机振荡,给出了合理的控制区间和运行通道。
The theory of linear regression and the theory of moving average are applied to analyse single data in time series, the model of a linear moving self regression forecast are given out.
应用推荐