Implementation of the model describe in the paper Random Walk in Feature-Sample Networks for Semi-supervised Classification.
I used sparse matrix in order to prevent memory error and optimize the algorithm. I also used this implementation to develop this paper