« home

Reinforcement Learning Greedy Policy

Creator: Petar Veličković (original)

A grid-world policy that greedily selects the locally best action under estimated state values, steering toward the goal while avoiding a penalty state.


Reinforcement Learning Greedy Policy

Download

PNGPNG (HD)PDFSVG

Code

reinforcement-learning-greedy-policy.typ (155 lines)