[danharaj]

A3C for now just to replicate the results in the paper. After that, I want to investigate more geometrically motivated methods like manifold learning and see if they can be adapted to processes.