Programming Exercise 6: Support Vector Machines

1.2.1 Gaussian Kernel 

高斯核求解公式 

sim=exp(-(x1-x2)'*(x1-x2)/2/sigma/sigma)

1.2.3 Example Dataset 3 

error=zeros(8,8);
options=[ 0.01, 0.03, 0.1, 0.3, 1, 3, 10, 30];
for i = 1:8
    for j =1:8
        C=options(i);
        sigma=options(j);
        %用训练集和新的C，sigmoid训练模型
        model= svmTrain(X, y, C, @(x1, x2) gaussianKernel(x1, x2, sigma));
        %预测
        predictions = svmPredict(model, Xval);
        %求误差
        error(i,j)=mean(double(predictions ~= yval));
    end
end
minerror=min(min(error));

[C_index,sigma_index]=find(error==minerror);
C=options(C_index);
sigma=options(sigma_index);

2.1.1 Vocabulary List

这一部分从vocab.txt文件中找到样例垃圾邮件中各单词的下标

    len_voca=length(vocabList);
    for i =1:len_voca 
        str_vocab=vocabList{i};
        if strcmp(str,str_vocab)==1
            word_indices=[word_indices,i];
            break;
        end
    end

2.2 Extracting Features from Emails

 

%将vacab中的单词提取出来
vocabList = getVocabList();
len_voca=length(vocabList);
for i =1:len_voca 
    %判断邮件中有没有这个单词
    if ismember(i,word_indices)
        x(i)=1;
    else
        x(i)=0;
    end
end