neuralnetwork_bp_classification.m

BP 神经网络 分类应用 简单实用

% BP 神经网络用于模式分类
% 使用平台 - Matlab6.5
% 作者：陆振波，海军工程大学
% 欢迎同行来信交流与合作，更多文章与程序下载请访问我的个人主页
% 电子邮件：[e-mail]
% 个人主页：http://luzhenbo.88uu.com.cn

clc
clear
close all

%---------------------------------------------------
% 产生训练样本与测试样本，每一列为一个样本

P1 = [rand(3,5),rand(3,5)+1,rand(3,5)+2];
T1 = [repmat([1;0;0],1,5),repmat([0;1;0],1,5),repmat([0;0;1],1,5)];    
%                      
% B =repmat Replicate and tile an array SyntaxB = repmat(A,m,n)
% B = repmat(A,[m n])
% B = repmat(A,[m n p...])
% repmat(A,m,n)
% DescriptionB = repmat(A,m,n) creates a large matrix B consisting of an m-by-n tiling of copies of A. The statement repmat(A,n) creates an n-by-n tiling. B = repmat(A,[m n]) accomplishes the same result as repmat(A,m,n). B = repmat(A,[m n p...]) produces a multidimensional (m-by-n-by-p-by-...) array composed of copies of A. A may be multidimensional. repmat(A,m,n) when A is a scalar, produces an m-by-n matrix filled with A's value. This can be much faster than a*ones(m,n) when m or n is large. ExamplesIn this example, repmat replicates 12 copies of the second-order identity matrix, resulting in a "checkerboard" pattern. B = repmat(eye(2),3,4)

% B =
%      1     0     1     0     1     0     1     0
%      0     1     0     1     0     1     0     1
%      1     0     1     0     1     0     1     0
%      0     1     0     1     0     1     0     1
%      1     0     1     0     1     0     1     0
%      0     1     0     1     0     1     0     1 
% The statement N = repmat(NaN,[2 3]) creates a 2-by-3 matrix of NaNs.
%      1     0     1     0     1     0     1     0
%      0     1     0     1     0     1     0     1
%      1     0     1     0     1     0     1     0
%      0     1     0     1     0     1     0     1
%      1     0     1     0     1     0     1     0
%      0     1     0     1     0     1     0     1 
% The statement N = repmat(NaN,[2 3]) creates a 2-by-3 matrix of NaNs.

P2 = [rand(3,5),rand(3,5)+1,rand(3,5)+2];
T2 = [repmat([1;0;0],1,5),repmat([0;1;0],1,5),repmat([0;0;1],1,5)];

%---------------------------------------------------
% 归一化

[PN1,minp,maxp] = premnmx(P1);
PN2 = tramnmx(P2,minp,maxp);

%---------------------------------------------------
% 设置网络参数

NodeNum = 10;                   % 隐层节点数 
TypeNum = 3;                    % 输出维数

TF1 = 'tansig';TF2 = 'purelin'; % 判别函数(缺省值)
%TF1 = 'tansig';TF2 = 'logsig';
%TF1 = 'logsig';TF2 = 'purelin';
%TF1 = 'tansig';TF2 = 'tansig';
%TF1 = 'logsig';TF2 = 'logsig';
%TF1 = 'purelin';TF2 = 'purelin';

net = newff(minmax(PN1),[NodeNum TypeNum],{TF1 TF2});

%---------------------------------------------------
% 指定训练参数

% net.trainFcn = 'traingd';  % 梯度下降算法
% net.trainFcn = 'traingdm'; % 动量梯度下降算法
% 
% net.trainFcn = 'traingda'; % 变学习率梯度下降算法
% net.trainFcn = 'traingdx'; % 变学习率动量梯度下降算法 
%
% (大型网络的首选算法 - 模式识别)
% net.trainFcn = 'trainrp';  % RPROP(弹性BP)算法,内存需求最小 
% 
% 共轭梯度算法 
% net.trainFcn = 'traincgf'; % Fletcher-Reeves修正算法
% net.trainFcn = 'traincgp'; % Polak-Ribiere修正算法,内存需求比Fletcher-Reeves修正算法略大
% net.trainFcn = 'traincgb'; % Powell-Beal复位算法,内存需求比Polak-Ribiere修正算法略大
% (大型网络的首选算法 - 函数拟合,模式识别)
% net.trainFcn = 'trainscg'; % Scaled Conjugate Gradient算法,内存需求与Fletcher-Reeves修正算法相同,计算量比上面三种算法都小很多 
% 
% net.trainFcn = 'trainbfg'; % Quasi-Newton Algorithms - BFGS Algorithm,计算量和内存需求均比共轭梯度算法大,但收敛比较快
% net.trainFcn = 'trainoss'; % One Step Secant Algorithm,计算量和内存需求均比BFGS算法小,比共轭梯度算法略大
% 
% (中小型网络的首选算法 - 函数拟合,模式识别)
net.trainFcn = 'trainlm';  % Levenberg-Marquardt算法,内存需求最大,收敛速度最快
%
% net.trainFcn = 'trainbr';  % 贝叶斯正则化算法
%
% 有代表性的五种算法为:'traingdx','trainrp','trainscg','trainoss', 'trainlm'

%---------------------%

net.trainParam.show = 1;          % 训练显示间隔
net.trainParam.lr = 0.3;            % 学习步长 - traingd,traingdm
net.trainParam.mc = 0.95;           % 动量项系数 - traingdm,traingdx
net.trainParam.mem_reduc = 10;      % 分块计算Hessian矩阵(仅对Levenberg-Marquardt算法有效)
net.trainParam.epochs = 1000;       % 最大训练次数
net.trainParam.goal = 1e-8;         % 最小均方误差
net.trainParam.min_grad = 1e-20;    % 最小梯度
net.trainParam.time = inf;          % 最大训练时间

%---------------------------------------------------
% 训练与测试

net = train(net,PN1,T1);             % 训练

%---------------------------------------------------
% 测试

Y1 = sim(net,PN1);             % 训练样本实际输出
Y2 = sim(net,PN2);             % 测试样本实际输出

Y1 = full(compet(Y1));         % 竞争输出
Y2 = full(compet(Y2));     




% View code for competDefault Topics  compet Competitive transfer function.
%    
%    Syntax
%  
%      A = compet(N)
%      info = compet(code)
%  
%    Description
%    
%      compet is a transfer function.  Transfer functions
%      calculate a layer's output from its net input.
%    
%      compet(N) takes one input argument,
%        N - SxQ matrix of net input (column) vectors.
%      and returns output vectors with 1 where each net input
%      vector has its maximum value, and 0 elsewhere.
%    
%      compet(code) returns information about this function.
%      These codes are defined:
%        'deriv'  - Name of derivative function.
%        'name'   - Full name.
%        'output' - Output range.
%        'active' - Active input range.
%  
%      compet does not have a derivative function.
%    
%    Examples
%  
%      Here we define a net input vector N, calculate the output,
%      and plot both with bar graphs.
%  
%        n = [0; 1; -0.5; 0.5];
%        a = compet(n);
%        subplot(2,1,1), bar(n), ylabel('n')
%        subplot(2,1,2), bar(a), ylabel('a')

%---------------------------------------------------
% 结果统计

Result = ~sum(abs(T1-Y1))                  % 正确分类显示为1
Percent1 = sum(Result)/length(Result)      % 训练样本正确分类率

Result = ~sum(abs(T2-Y2))                  % 正确分类显示为1
Percent2 = sum(Result)/length(Result)      % 测试样本正确分类率