Deep Reinforcement Learning with Custom Environment

This project involves creating a custom Blackjack environment and training an AI using reinforcement learning techniques, specifically Proximal Policy Optimization (PPO) and Deep Q-Network (DQN). The goal is to teach the AI to play Blackjack and achieve the best possible win rate.

Tutorial for using Stable Baselines 3 for creating custom policies

Custom-built Proximal Policy Optimization (PPO) agent learns to master a 2D shooter game. Features from-scratch PPO implementation, Pygame-based environment, and OpenAI Gym integration. Showcases reinforcement learning in game AI, combining advanced algorithm development with practical game design.

Repository for a custom OpenAI Gym compatible environment for the Parrot Drone ANAFI 4K.

The program uses the DDPG algorithm and tf_agents library to train an agent in a custom environment called "TargetSeeker"

My Debian configuration dotfiles (backup). Minimalistic environment and instruments with old-school fonts.

Reinforcement Learning and Deeep reinforcement Learning

This repository contains implementations for reward shaping based governance kernel layer experiments

OpenAI's PPO baseline applied to the classic game of Snake

We provide the code repository for our paper This repository includes the necessary code to replicate our experiments and utilize our DRL model for spacecraft trajectory planning. By accessing the repository, researchers and practitioners can benefit from our approach to efficiently transfer spacecraft to GEO using low-thrust propulsion systems.