This repository contains an PyTorch implementation of a Deep Reinforcement Learning approach to Financial Portfolio Management. It is motivated by A Deep Reinforcement Learning Framework for the Financial Portfolio Management Problem by Jiang et. al. 2017. It is additionally motivated by a similar TensorFlow implementation which can be found here
In this work we implement two deep reinforcement learning algorithms, Deep Deterministic Policy Gradient (DDPG) and Proximal Policy Optimization (PPO) to allocate portions of a fund to different financial products with the goal of maximising returns. Additionally we aim to investigate the use of Transformers for the neutral network structures.