calculate_error.m

来自「最新的模式识别分类工具箱,希望对朋友们有用!」· M 代码 · 共 20 行

function [train_err, test_err] = calculate_error (D, train_features, train_targets, test_features, test_targets, region, Nclasses)

% Calculate error (used by the main calculation functions)

train_err   = zeros(Nclasses+1,1);
test_err    = zeros(Nclasses+1,1);

[classify, err]  = classification_error(D, train_features, train_targets, region);
for j = 1:Nclasses,
   train_err(j)   = 1 - classify(j,j);    
end
train_err(Nclasses+1) = err;     

[classify, err]  = classification_error(D, test_features, test_targets, region);
for j = 1:Nclasses,
   test_err(j)   = 1 - classify(j,j);    
end
test_err(Nclasses+1) = err;