The aim of this thesis is to explore the potential of a Deep Reinforcement Learning approach to the Portfolio Optimization problem. Four different types of Reinforcement Learning algorithms – Advantage Actor-Critic (A2C), ...