Reinforcement Learning applied to Automatic Investiment on the Brazilian Stock Market

• Grant number: 24/04827-1
• Support Opportunities: Scholarships in Brazil - Scientific Initiation
• Start date: June 01, 2024
• End date: May 31, 2025
• Field of knowledge: Physical Sciences and Mathematics - Computer Science - Computing Methodologies and Techniques
• Principal Investigator: André Carlos Ponce de Leon Ferreira de Carvalho
• Grantee: Paulo Ricardo Sturion
• Host Institution: Instituto de Ciências Matemáticas e de Computação (ICMC). Universidade de São Paulo (USP). São Carlos , SP, Brazil

Abstract

The project aims to develop different Reinforcement Learning techniques (Proximal Policy Optimization, Recurrent Reinforcement Learning and Deep-Q-Learning) in a comparative way for the application to the problem of automatic investment on the Brazilian stock market - specifically for daily operations (swing trade ). Therefore, it is proposed: the development of a database with historical data of different attributes regarding the main stocks on the market; study and implementation of the chosen Reinforcement Learning methods; perform simulations and experiments using the collected data; assessment and comparison of the different methods investigated by means of well established selected metrics; and the analysis of the results and publication of articles.