Wavelet decomposition problems have been modeled as linear programs – but only as extremely dense problems. Both revised simplex and interior point methods have difficulty with dense linear programs. The question then is how to get around that issue. In our experiments the standard method outperforms a revised implementation for these problems. Moreover, the standard method can be easily and scalably distributed. Hence the standard simplex method should be useful in solving wavelet decomposition problems.
View WAVELET DECOMPOSITION VIA THE STANDARD TABLEAU SIMPLEX METHOD OF LINEAR PROGRAMMING