dyna-algorithm-learns-model-generates-simulated-transitions

IN premise — entries/2026/06/21/wiki-Reinforcement_learning-chunk-2.md

Created 2026-06-21T09:55:52+00:00

The Dyna algorithm learns a model from real experience and generates simulated transitions to supplement real ones for value function updates, improving sample efficiency