Financial Portfolio Management using Reinforcement Learning

  • Jash Patani, Saurav Nair, Kalpesh Mehta, Ameya Sankhe, Pratik Kanani

Abstract

Portfolio management is generally defined as the science of choosing the right investment policies and making optimal investment decisions with the aim of minimizing risk and maximizing returns. Speaking in terms of layman’s words, the technique of managing an individual’s investment is known as portfolio management. Portfolio management requires suggesting the best investment plans to any company, organization or individual based on various parameters like income, budget, risk aversion, and many more such factors. Portfolio management minimizes the risk involved and maximizes the gains. In the world of finance, even minor advantages can be multiplied to give exponential gains. As Artificial Intelligence is evolving, its uses in different fields are coming in the picture. Currently, it is also being used in the field of Finance. It is still not perfect, but it is slowly evolving. If it were easy everyone would have trained a model and perform inferences to be rich. In this paper, we have used Reinforcement Learning for Portfolio management. Our experiments are performed on data sets of Indian stocks.

Keywords: Reinforcement Learning, Portfolio Management, DQN, T-DQN, Double-DQN

Published
2020-05-30
How to Cite
Jash Patani, Saurav Nair, Kalpesh Mehta, Ameya Sankhe, Pratik Kanani. (2020). Financial Portfolio Management using Reinforcement Learning. International Journal of Advanced Science and Technology, 29(05), 9740 - 9751. Retrieved from http://sersc.org/journals/index.php/IJAST/article/view/19442