Somasse, Gbetonmasse Blaise
Developed an autonomous stock trading algorithm. Each group member created their own trading system, and parts of each system were combined into a single system. The final system consists of a separate algorithm for selecting stocks, buying stocks, allocation money, and selling stocks. The buy process uses a Random Forest machine learning algorithm trained on historical stock prices and modified bars based on volume to predict when to buy. The system and backtesting software were developed in Python using modules Numpy, Pandas, Scikit-learn, and Matplotlib. The system was tested across short, intermediate and long term periods. The average annualized return on investment for the three time periods was 26.67%.
Worcester Polytechnic Institute
Interactive Qualifying Project
All authors have granted to WPI a nonexclusive royalty-free license to distribute copies of the work, subject to other agreements. Copyright is held by the author or authors, with all rights reserved, unless otherwise noted.