%% LVQ神经网络的分类——乳腺肿瘤诊断
%% 清空环境变量
clear all
clc
warning off
%% 导入数据
load data.mat
a=randperm(569);
Train=data(a(1:500),:);
Test=data(a(501:end),:);
% 训练数据
P_train=Train(:,3:end)';
Tc_train=Train(:,2)';
T_train=ind2vec(Tc_train);
% 测试数据
P_test=Test(:,3:end)';
Tc_test=Test(:,2)';
%% K-fold交叉验证确定最佳神经元个数
k_fold=5;
Indices=crossvalind('Kfold',size(P_train,2),k_fold);
error_min=10e10;
best_number=1;
best_input=[];
best_output=[];
best_train_set_index=[];
best_validation_set_index=[];
h=waitbar(0,'正在寻找最佳神经元个数.....');
for i=1:k_fold
% 验证集标号
validation_set_index=(Indices==i);
% 训练集标号
train_set_index=~validation_set_index;
% 验证集
validation_set_input=P_train(:,validation_set_index);
validation_set_output=T_train(:,validation_set_index);
% 训练集
train_set_input=P_train(:,train_set_index);
train_set_output=T_train(:,train_set_index);
for number=10:30
count_B_train=length(find(Tc_train(:,train_set_index)==1));
count_M_train=length(find(Tc_train(:,train_set_index)==2));
rate_B=count_B_train/length(find(train_set_index==1));
rate_M=count_M_train/length(find(train_set_index==1));
net=newlvq(minmax(train_set_input),number,[rate_B rate_M]);
% 设置网络参数
net.trainParam.epochs=1000;
net.trainParam.show=10;
net.trainParam.lr=0.1;
net.trainParam.goal=0.1;
% 训练网络
net=train(net,train_set_input,train_set_output);
waitbar(((i-1)*21+number)/114,h);
%% 仿真测试
T_sim=sim(net,validation_set_input);
Tc_sim=vec2ind(T_sim);
error=length(find(Tc_sim~=Tc_train(:,validation_set_index)));
if error<error_min
error_min=error;
best_number=number;
best_input=train_set_input;
best_output=train_set_output;
best_train_set_index=train_set_index;
best_validation_set_index=validation_set_index;
end
end
end
disp(['经过交叉验证，得到的最佳神经元个数为：' num2str(best_number)]);
close(h);
%% 创建网络
count_B_train=length(find(Tc_train(:,best_train_set_index)==1));
count_M_train=length(find(Tc_train(:,best_train_set_index)==2));
rate_B=count_B_train/length(find(train_set_index==1));
rate_M=count_M_train/length(find(train_set_index==1));
net=newlvq(minmax(best_input),best_number,[rate_B rate_M]);
% 设置网络参数
net.trainParam.epochs=1000;
net.trainParam.show=10;
net.trainParam.lr=0.1;
net.trainParam.goal=0.1;
%% 训练网络
net=train(net,best_input,best_output);
%% 仿真测试
T_sim=sim(net,P_test);
Tc_sim=vec2ind(T_sim);
result=[Tc_sim;Tc_test]
%% 结果显示
total_B=length(find(data(:,2)==1));
total_M=length(find(data(:,2)==2));
count_B_validation=length(find(Tc_train(:,best_validation_set_index)==1));
count_M_validation=length(find(Tc_train(:,best_validation_set_index)==2));
number_B=length(find(Tc_test==1));
number_M=length(find(Tc_test==2));
number_B_sim=length(find(Tc_sim==1 & Tc_test==1));
number_M_sim=length(find(Tc_sim==2 &Tc_test==2));
disp(['病例总数：' num2str(569)...
' 良性：' num2str(total_B)...
' 恶性：' num2str(total_M)]);
disp(['训练集病例总数：' num2str(length(find(best_train_set_index==1)))...
' 良性：' num2str(count_B_train)...
' 恶性：' num2str(count_M_train)]);
disp(['验证集病例总数：' num2str(length(find(best_validation_set_index==1)))...
' 良性：' num2str(count_B_validation)...
' 恶性：' num2str(count_M_validation)]);
disp(['测试集病例总数：' num2str(69)...
' 良性：' num2str(number_B)...
' 恶性：' num2str(number_M)]);
disp(['良性乳腺肿瘤确诊：' num2str(number_B_sim)...
' 误诊：' num2str(number_B-number_B_sim)...
' 确诊率p1=' num2str(number_B_sim/number_B*100) '%']);
disp(['恶性乳腺肿瘤确诊：' num2str(number_M_sim)...
' 误诊：' num2str(number_M-number_M_sim)...
' 确诊率p2=' num2str(number_M_sim/number_M*100) '%']);
