Here is a great introduction to PCA for beginners and I can't do better than this Princeton PCA
After reading this I was a bit confused on how to apply this in my matlab code. Let's go through it
Here md contains my data, suppose that is a matrix of size 10000 x 15 . Now generally I should do some analysis on the variances of eigenvectors before selecting the final dimension that I want. But let's just say that I want 10 .
eigenvectors is a 15 x 15 matrix whose columns are my eigenvectors. Now I project my original data onto the space given by this and get the reduced matrix in md.
After reading this I was a bit confused on how to apply this in my matlab code. Let's go through it
[eigenvectors,score,latent] = pca(md);
md = md * eigenvectors(:,1:10);
fprintf('Eigenvalues for the data \n');
disp(latent);
Here md contains my data, suppose that is a matrix of size 10000 x 15 . Now generally I should do some analysis on the variances of eigenvectors before selecting the final dimension that I want. But let's just say that I want 10 .
eigenvectors is a 15 x 15 matrix whose columns are my eigenvectors. Now I project my original data onto the space given by this and get the reduced matrix in md.
