摘要
针对自变量与因变量很少完全服从线性模型的实际,并克服常规最小二乘法处理曲线拟合中忽视自变量的误差的不足,提出总体最小二乘法拟合曲线,同时顾及了自变量和因变量的误差,提高了拟合的可靠性。算例结果表明,总体最小二乘法曲线拟合效果优于常规最小二乘法。
According to independent variables actual rarely obey linear model, and overcome the shortage of conventional least squares ignoring the error of independent variables to process curve fitting. The authors put forward total least squares fitting curve, which takes into account the errors of independent and dependent variables, and improves the reliability of the fitting. The example shows that total least squares curve fitting result is better than conventional least squares.
出处
《地矿测绘》
2012年第3期4-6,共3页
Surveying and Mapping of Geology and Mineral Resources
关键词
总体最小二乘法
自变量
曲线拟合
最小二乘法
精度
total least squares method
independent variables
curve fitting
least squares method
precision
