Question

作为关于机器学习问题的基本问题，我想知道如何使用此描述来计算分类器的泛化：

数据集属性：
- 样本数：220
- 功能数量：19
- 班级数：7
分类方法：
- 支持矢量机。
- 就像你现在一样，我的分类任务与将一个样本分为7个类有关，所以我从一对一的方法中使用，然后利用多数投票的方法。 [分类程序：我创建了21个分类器，然后对于每个样本我分别用21个分类器进行分类。直到这一步，我得到了21个类别标签（1,2,3,4,5,6,7）。然后，我进入下一步并计算每个标签的数量（这称为多数投票），例如输入x：#Class Label ONE = 7，＃TWO = 5，＃THREE = 4，＃FOUR = 1， #FIVE = 1，＃SIX = 1，＃SEVEN = 2.]。我会将每个样本分配给在Majority-Process中具有更高学位的任何课程。

因此，使用此说明，如何分析我的分类器重新归结为泛化问题。

我欢迎任何可以帮助我的MATLAB代码。

load('dataset.mat');

X_train = data_train;
X_test = data_test;

[train_size, feature_size, class_size] = size(X_train);
[test_size] = size(X_test,1);

x = [];
for index=1:class_size
x=[x;X_test(:,:,index)];
end
X_test = x;

correct_label = [];
for index=1:class_size
     correct_label = [correct_label;index*ones(test_size,1)];
end

g = [];
C = 1;
for i=class_size:-1:1
    for j=1:(i-1)
        train = [X_train(:,:,j);X_train(:,:,i)];
        y = [j*ones(train_size,1);i*ones(train_size,1)];      
        svmModel = fitcsvm(train, y,'Standardize',true,'KernelFunction','GHI_Kernel');
        group = predict(svmModel, X_test);
        g = [g,group];
    end
end
temp = zeros(size(X_test,1),class_size);
n = size(X_test,1);
for i=1:n
     for j=1:class_size
          temp(i,j) = sum(((g(i,:)==j)'));
     end
end
%% esm1 contains the predicted class label
%% val1 contains the number of voting for predicted class label
[val1, esm1] = max(temp,[],2);

%% Calculating Confidence Matrix
for idx=1:n
     temp(idx,esm1(idx))=0;
end
[val2, esm2] = max(temp,[],2);
confidence_matrix = zeros(class_size, class_size);
counter = zeros(class_size, class_size);

for idx=1:n
     confidence_matrix(esm1(idx), correct_label(idx))...
          = confidence_matrix(esm1(idx), correct_label(idx)) + ...
          (val1(idx) - val2(idx))/val1(idx);
     counter(esm1(idx), correct_label(idx)) = ...
          counter(esm1(idx), correct_label(idx)) + 1;
end

for idx=1:class_size
    confidence_matrix(:,idx) = confidence_matrix(:,idx)/sum(confidence_matrix(:,idx));
end

CCR = sum(esm1 == correct_label)/size(correct_label,1);
plot_confusion_matrix(esm1, correct_label);
error = 1 - CCR;
confusion_matrix = confusionmat(esm1, correct_label);

如何用离开交叉验证来计算一对一分类器的推广？

0 个答案: