Reinforcement learning research framework implementing Double/Dueling DQN with PER, N-step Bellman returns, and Dyna-Q planning in pure NumPy. Investigates convergence properties, optimization stability, and structured high-dimensional state representations.
reinforcement-learning deep-reinforcement-learning q-learning self-learning autonomous-agents cognitive-architecture mental-map neuromorphic-learning
-
Updated
Apr 23, 2026 - Python