[dw,ls] = (W,P,Z,N,A,T, E,gW,gA,D,LP,LS)
其参数意义与中的参数意义相同，只是权值调整的方法不同,此处不再赘述 。
4. 实现
利用神经网络工具箱提供的函数可以方便地在环境下实现上述步骤 。
%% LVQ神经网络的分类——乳腺肿瘤诊断%% 清空环境变量clear allclcwarning off%% 导入数据load data.mata = 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)';%% 创建网络count_B = length(find(Tc_train == 1));count_M = length(find(Tc_train == 2));rate_B = count_B/500;rate_M = count_M/500;net = newlvq(minmax(P_train),20,[rate_B rate_M],0.01,'learnlv1');% 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,P_train,T_train);%% 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));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(500)...'良性：' num2str(count_B)...'恶性：' num2str(count_M)]);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) '%']);
某次运行神经网络训练测试的结果如下:
病例总数:569良性:357恶性 :212训练集病例总数:500良性:312恶性 :188测试集病例总数:69良性:45恶性:24良性乳腺肿瘤确诊:43误诊:2确诊率 p1= 95.5556%恶性乳腺肿瘤确诊:20误诊:4确诊率 p2= 83.3333%
从上述结果可以看出，在69组测试集数据中,有6组数据误诊断(2组将良性乳腺肿瘤误诊为恶性乳腺肿瘤,4组将恶性乳腺肿瘤误诊为良性乳腺肿瘤),平均诊断正确率达91. 3%(63/69) 。实验结果表明,将LVQ神经网络应用于模式识别是可行的 。
5.案例拓展
本小节将对比上述 LVQ神经网络经网络两 种模 型的效果，BP神经网络模型的程序如下：
%% BP神经网络的分类——乳腺肿瘤诊断%% 创建网络net = newff(minmax(P_train),[50 1],{'tansig','purelin'},'trainlm');%% 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,P_train,Tc_train);%% 仿真测试T_sim = sim(net,P_test);for i = 1:length(T_sim)if T_sim(i) <= 1.5T_sim(i) = 1;elseT_sim(i) = 2;endendresult = [T_sim;Tc_test]number_B = length(find(Tc_test == 1));number_M = length(find(Tc_test == 2));number_B_sim = length(find(T_sim == 1 & Tc_test == 1));number_M_sim = length(find(T_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(500)...'良性：' num2str(count_B)...'恶性：' num2str(count_M)]);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) '%']);