基于模糊点数据的线性回归模型在判别分析中的应用

The Application of Linear Regression Model to Discriminant Analysis Based on Fuzzy Points Data

在经典回归分析理论中,训练数据在回归方程的构建中是被同等对待的.然而,在许多实际问题中,训练数据的作用是不同的,通常有些训练数据比其他数据更重要.为此,给每个训练数据赋予一个置信权重(这里的样本称为模糊点样本数据),并且给出了确定该置信权重的几种常用方法,讨论了基于模糊点数据的线性回归模型用于判别分析的情况.最后给出了数值例子.

For classical regression analysis, all training data are treated uniformly in the construction of regression equation. However, in many real-world problems, the effects of the training data to the fitting curve may be different. It is often that some training data may be more important than others. In this paper, a fuzzy membership degree is applied to each training data and several possible cases are given. Samples in here are called fuzzy points sample data. The application of the linear regression model to discriminant analysis based on fuzzy points data is discussed. Finally, two running examples are presented.