I have tried to ask the same question in the math site on stack exchange but no real joy. Apologises I'd this is the wrong forum.
I am trying to compute the PCA of a spectrogram.. It is 486x128 after I compute the covariance matrix, the resulting matrix is a square matrix containing 128x128 dimensions. Now, I want to compute the Eigen values for this. So far, I have computed the eigen values for a 2x2, 3x3 as well as a 4x4 matrix and this makes sense.
I cannot for the life of me think how I would take the concepts from this and apply it to a 128x128 matrix. When I compute the Eigenvalues and vectors in Matlab, returned is a similar or same size 128x128 matrix containing the eigenvalues and eigenvectors.
In order to solve this problem, shall I therefore compute the eigen value for each of the blocks contained in the spectrogram (I.e. this will give me 128 eigen values) or is there another way to calculate based off this size of a matrix?
I hope someone can help