🏓Deep learning model is presented to successfully learn control policies directly from high-dimensional sensory input using reinforcement learning. The model is a convolutional neural network, trained with a variant of Q-learning, whose input is raw pixels and whose output is a value function estimating future rewards in RL Pong environment.

A mini project during 3 days of Tet Holiday 2023

[TNNLS] PGDQN: A generalized and efficient preference-guided epsilon-greedy policy equipped DQN for Atari and Autonomous Driving

A Deep Q-Network reinforcement learning model trained to safely land a lunar lander in the Farama Gymnasium

This project implements Value Iteration and Q-Learning algorithms to solve a variety of gridworld mazes and puzzles. It provides pre-defined policies that can be customized by adjusting parameters and policy optimization through iterative reinforcement learning. It also brings exploration capabilities to the agent with Epsilon Greedy Q-Learning.