I need step by step neat and clean handwritten solution in A 4 page

otherwise i will do downvote !!

Problem 2 [5 points) Consider a data matrix of dimension mx n. In some applications the role of points and dimensions can be interchanged. For example, given a document corpus represented as a matrix of type "documents x words", we may want to analyze documents based on which words occur in them, or we may want to analyze words based on which documents they appear in. So it is meaningful to perform PCA both with respect to the rows of a matrix and with respect to its columns. As we discussed in the video lectures, PCA relies on SVD. Moreover, since (UEVT)' = VETUT = VE'U", where x differs from only in terms of size, performing SVD on a matrix gives also the SVD on its transpose. Does this argument imply that a single SVD operation is sufficient to perform PCA both on the rows and the columns of a data matrix? Justify your