My Learnings in ML

Here is some really cute 3 lines of matlab code to sample from non i.i.d data. Here it is - data = rand(n,1); % sample from i.i.d data [U S V] = svd(C); % C represents the covariance matrix of the data z = U*sqrt(S)*data; In order to understand why it works, consider the innermost term of the Gaussian distribution which is . (x-M)'inv(C)(x -M) = (x -M)'sqrt(inv(C))(x -M) = (x -M)'sqrt(inv(U*S*inv(U)))(x -M) %V' = inv(U) as C is symmetric Solving it further, we get the required answer.

Using this blog, I want to talk about my learning in Machine Learning. I am an eternal student in the area and the vast expanse of algorithms in the field continue to amaze me. Via this blog, I will try to compile my list of fascinating algorithms.