Awesome

spls

Sparse Partial Least Squares method using L1 and L2 norm constraints. Please check the paper by Monteiro et al. 2016 for details.

The computation of the first weight vector pair (u,v) should be done using spls.m. Then, matrix deflation should be performed using proj_def.m, and the process repeated in order to find the next weight vector pair.