Tag: Policy Gradient
-
Part 3: Policy Gradient vs. Q-Learning: Choosing the Right Agent
DQN vs PPO vs SAC: when each algorithm wins. Math foundations, code examples, and a decision tree for picking the right RL agent for your task.
-
Policy Gradient ์๋ฒฝ ๊ฐ์ด๋: REINFORCE๋ถํฐ PPO๊น์ง ์ํ ์๋ฆฌ์ PyTorch ๊ตฌํ
Policy Gradient from scratch: REINFORCE to PPO with PyTorch code. Reduce variance 100x using baselines and see why PPO beats vanilla PG.
-
PPO and A3C for Complex Game AI Environments
Why PPO replaced DQN for game AI: actor-critic variance reduction, continuous control, and 10x faster convergence on Atari benchmarks.