曲线回归参数估计的改进与回归预测模型的校正  

THE IMPROVEMENT OF PARAMETER ESTIMATION OF CURVE REGRESSION AND THE RECTIFICATION OF REGRESSION FORECASTING MODEL

作　　者：仲莉娜 申花实[1] 王胜美[1] 

机构地区：[1]东北林业大学

出　　处：《东北林业大学学报》1995年第1期141-145,共5页Journal of Northeast Forestry University

摘　　要：为了使回归预测模型更具有实用性，采用渐近最小二来法对指数函数、幂函数一类回归曲线的参数估计做了改进。在增加新的样本单元资料时，用增长记忆递推公式对原始回归预测模型进行了校正，得到新的预测模型。In order to make regression forecasting model be more useful, this paper improved the parameter estimation of regression curves of exponential functions and power functions by using the improved asymptotic least square method. The original regression forecasting model was rectified by using recursion formula of increasing memory. When the sample units increased, a new forecasting model was got.

关 键 词：回归方程 参数估计 最小二乘法 

分 类 号：O212.1[理学—概率论与数理统计]