A sophisticated portfolio allocation system that compares traditional Mean-Variance Optimization (MVO) with advanced Deep Reinforcement Learning (RL) approaches, and adds inverse optimisation for risk aversion.

Use DeepDow to optimise portfolio using Pytorch

Code for Chapter 5 of the thesis: 'Portfolio Optimisation in a Fractional Stochastic Environment with Deep Reinforcement Learning Applications'.