I have been studying the concept of PCA and its implementation for dimensionality reduction for more than 1 month. My goal is to classify a hyperspectral image using sparse representation by the linear combination concept which is as follow:

$$y = Dx$$

So consider $D$ as a dictionary with $d\times B$ dimension where $d=3000$ is the number of samples and $B=200$ is the number of band/channel.

Now I am trying to construct the $D$ by this mean that the classes (sub-dictionaries) are well separated. Therefore I want to apply PCA to individual sub-dictionary in order to form the main dictionary.

However, my goal is to apply PCA on hyperspectral satellite imagery like this.

I have implemented the PCA in Octave and project my data on that particular low dimension. But my question is should I reduce the number of training pixels(observation=d$$) or reduce the variable dimension ($B$)?

Since I use sparse representation and dictionary concept then reducing the dimension of $d$ (observation pixels for individual classes) is more make sense rather than reducing the number of features ($B$). But I am not sure if I am right or not.

After constructing the $D$ should I transform $y$ to PCA dimension before computing its spars coefficients or not?

  • $\begingroup$ typically, PCA selects the principal components from your multi-dimensional signal, i.e. in your case, it reduces $B$ without reducing the number of observation you've got. $\endgroup$ Commented Jan 11, 2019 at 16:03
  • $\begingroup$ But I manipulate the covariance matrix to have a pairwise pixels evaluation by means of covariance quantity, instead of computing covariance between variables. therefore I am able to compute the eigenvectors and their corresponding eigenvalues for samples (we have lots of reduction vector [pixel]) $\endgroup$
    – morteza
    Commented Jan 14, 2019 at 1:20
  • $\begingroup$ that can be presented by A scalar multiplication of another. So lets only use certainty, not uncertainty (in terms of variance of the data). $\endgroup$
    – morteza
    Commented Jan 14, 2019 at 1:29
  • $\begingroup$ The linear algebra five us two amazing concept cal Linearly-Dependent and Linearly-Independent. That is the way we can cover the main info with only the main numbers. $\endgroup$
    – morteza
    Commented Jan 14, 2019 at 1:30


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Browse other questions tagged or ask your own question.