MOOC Machine Learning 作业交流帖7

wk7 K-means Clustering and Principal Component Analysis

最近正在学习MOOC上的经典课程：Machine learning (by Andrew Ng)， 具体课程链接：MACHINE LEARNING
根据进度将作业的关键代码部分贴上，仅供交流与讨论。

• findClosestCentroids

len = zeros(K, 1); 
for i = 1:size(X,1)
    for j = 1:K 
        len(j) = norm(X(i, :) - centroids(j, :))^2; 
    end 
[~,idx(i)] = min(len); %% return the row of the min(len), that is return the exact cluster this training example belongs to
end 

• computeCentroids

for k = 1:K 
ind = find(idx == k); %% return the row number, that is the 'ith' training number
centroids(k, :) = mean(X(ind, :)); 
end

• pca

sigma=(X'*X)/m;
[U,S,~]=svd(sigma);

• projectData

Ureduced=U(:,1:K); %%%n*k
Z=X*Ureduced;      %%%X:m*n

• recoverData

X_rec = Z * U(:, 1:K)';

——转载请注明出处