The project is implemented in Python (Rmax-arch1_batch_limitedtraj.py) with the goal of finding the sample complexity of the M-FOS algorithm empirically.
Inner game Environment: Matching Pennies (env_mp_simple.py)
Meta-agent: Rmax (rmax_1_batch_limitedtraj.py)