REINFORCE Algorithm

An approach to computing gradients where stochastic nodes are involved. Originally introduced in the context of policy-based Reinforcement learning (RL).

Emacs 30.1.90 (Org mode 9.7.11)