Fisher判别法的MATLAB程序

就问你爱信不信

已于 2025-07-02 15:30:56 修改

阅读量6.5k

点赞数 3

CC 4.0 BY-SA版权

文章标签： matlab 算法

于 2020-11-18 13:47:30 首次发布

本文链接：https://blog.csdn.net/ibianma/article/details/109771797

这篇博客介绍了如何在MATLAB中应用Fisher判别法进行多元统计回归分析，展示了具体的实验结果。

摘要生成于 C知道，由 DeepSeek-R1 满血版支持，前往体验 >

多元统计回归分析

在这里插入图片描述

function [COEFF,DIST,CLASS] = dclass(X1,X2,X)
%    dclass作为两个模式类的距离判别分析
%    语法[COFEE,DIST,CLASS]=dclass(X1,X2,X)
%    x1,x2——分别为类1、类2的训练样本“样品×变量”矩阵
%    x——为待判样品的“样品×变量”矩阵
%    COEFF——判别函数的系数向量
%    CLASS——待判样品的分类
[N1,~]=size(X1);
[N2,~]=size(X2);
[N,~]=size(X);
MEANX1=mean(X1);
MEANX2=mean(X2);
COVX1=(N1-1)*cov(X1);
COVX2=(N2-1)*cov(X2);
MEAN=(MEANX1+MEANX2)./2;
COV=(COVX1+COVX2)./(N1+N2-2);
COEFF=inv(COV)*(MEANX1-MEANX2)';
DIST=[];
CLASS=[];
for byk=1:N
    w=(X(byk,:)-MEAN)*COEFF;
    if w>0
        r=1;
    else
        r=2;
    end
    DIST=[DIST,w];
    CLASS=[CLASS,r];
end
COEFF=COEFF';
end

实验结果

> clear all; clc; X1=[21.3  124.89 35.43 73.98 93.01  20.58 43.97 433.73
>     21.13 168.69 40.81 70.12 74.32  15.46 50.9  422.74
>     19.96 142.24 43.33 50.74 101.77 12.92 53.44 394.55]; X2=[21.5 122.39 29.08 51.64	55.04 11.3	54.88 288.13
>     18.25	90.21	24.45	32.44	62.48	7.45	47.5	178.84
>     21.84	66.38	18.05	31.32	74.48	8.19	34.97	177.45
>     21.37	67.08	20.28	35.27	81.07	10.94	39.46	182.2
>     22.74	115.88	28.21	42.44	58.07	9.63	48.65	194.85
>     20.22	88.94	18.54	35.63	65.72	8.81	50.29	186.52
>     21.33	75.5	14	    29.56	69.29	8.24	42.08	165.9
>     18.61	122.51	27.07	42.5	63.47	15.38	36.14	240.92
>     19.61	107.13	32.85	35.77	61.34	7.53	34.6	142.23
>     25.56	171.65	22.3	40.53	57.13	12.6	54.03	225.08
>     18.75	104.68	15.55	35.61	51.8	11.18	36.27	142.72
>     18.27	88.34	19.07	43.19	72.97	12.59	42.16	200.18
>     19.07	73.18	18.01	29.38	64.51	8.91	38.14	155.45
>     18.76	102.67	21.87	30.47	64.33	11.99	42.14	168.17
>     20.25	104.45	20.72	38.15	62.98	12.67	39.16	213.56
>     18.7	131.35	11.69	32.06	41.54	10.84	42.77	178.51
>     16.16	139.92	12.98	23.58	24.87	10.76	32.35	144.21
>     18.18	120.39	26.18	37.94	68.16	11.64	38.48	246.37
>     18.53	109.95	21.49	33.04	50.98	10.88	33.96	183.85
>     18.33	92.43	25.38	32.19	56.32	14	    38.57	144.82
>     22.3	99.08	33.36	32.01	52.06	7.04	32.85	190.04
>     20.03	70.75	19.75	34.95	53.29	10.55	38.2	189.41
>     18.68	72.74	23.72	38.69	62.41	9.65	35.26	170.12
>     20.33	75.64	20.88	33.85	53.81	10.06	32.82	171.32
>     19.75	70.24	18.67	36.71	61.75	10.08	40.26	165.22
>     21.03	78.55	14.35	34.33	64.98	9.83	33.87	161.67]; X=[ 23.68	173.30	17.43	43.59	53.66	16.86	65.02	385.94
>     29.67	146.90	64.51	54.36	86.10	14.77	32.19	193.10]; [COEFF,DIST,CLASS] = dclass(X1,X2,X)

CLASS =

 2     2 根据结果，广东和西藏被归为第二大类