Input from PCA to train in SVM

Hi Everyone. I had done PCA to reduce the feature but I am really confused what input from the PCA should i used to train in SVM.

I used this function :

[coeff,score,latent,tsquared,explained] = pca(fextracted);

I read somewhere they said the score should be the input. However, I dont think score is the correct one to be used since it has the same array size as my original data.

Answer 1

[the cyclist]

I have written a "tutorial" on how to use and interpret MATLAB's pca function here. Lots of users have asked questions (some of which are very similar to yours), and I have tried to answer them. I suggest you check that thread out, and see if it helps.

[nurin noor]

Hi Cyclist,

I've studied the thread and tried out to my code. It worked perfectly! Thank you so much!