Simulating Algorithmic Crypto Trading Strategies Using Reinforcement Learning

In this study, we investigate using the AMP (Adversarial Motion Priors) agent for cryptocurrency trading. Our main objective is to simulate trading algorithms using Reinforcement Learning and Generative Adversial Networks. Our goal is to refine trading decisions by simulating custom moving average crossover strategies. While our initial results show promise, our work is still in progress to increase stability and optimize decision making. Future research and improvements will involve subjecting the AMP agent to a variety of algorithmic crypto trading strategies to comprehensively validate and establish its effectiveness.