The SVD is a very useful tool for analysing a matrix, and the solutions to many problems can simply be read off from the decomposed matrices. For example, the null- space of the matrix (the solution to A x = 0 ) is simply the columns of Q 2 which correspond