A mass ssychological perspective on financial markets using meta reinforcement-learning

Abstract

Reinforcement learning (RL) is a computational framework for sequential decision-making, which combines control methods with machine-learning techniques, and is the state-of-the-art for solving large-scale decision problems. In this work, we pursue a novel approach to predicting the financial markets by analysing the market demography based on an individuals risk appetite. This is then used to predict the future trend of the particular product pricing. Based on previous studies, we extend the Temporal-Difference based Reinforcement Learning approach to represent individual agents with a particular risk appetite. We then train a meta-agent that is capable of switching between the optimal sections of the risk spectrum based on the decisions of the base agent.