...
Run first the routine XRM1_matrix_comp (this may take some time) and then XRM3_NoCrimeData_comp to create the matrix A and a noisy measurement mncn for the model Ax=mncn of x-ray tomography.
Now you should compute the TSVD solution of Ax=mncn by using M singular values (vectors). To do this, add necessary code after line 25 to the routine XRM6_truncSVD_comp.m; the code should compute the TSVD solution and save it to variable x. \ [Hint: use Matlab functions svd.m, size.m and basic matrix/vector operations.\] Wiki Markup
Having implemented the TSVD code to XRM6_truncSVD_comp.m
a) compute the TSVD solutions using 1, 10, 100, 500 and 700 singular vectors
b) what happens to the reconstruction with 700 singular vectors if the noiselevel (in XRM3_NoCrimeData_comp.m) is increased to 5% or to 10%?
...