Skip to content

AI in Finance: Machine Learning Models for Stock Price Predication and Auto-Trading

April 3, 2024 @
1:30 p.m.
- 2:30 p.m. Eastern Time
Loading Events
  • This event has ended.

Please join the Data Science Seminar this Wednesday for a talk by Dr. Timothy Li. Tim currently serves as principal data scientist and vice president at the Central Modeling Team of Citizens Bank. He will present two of his earlier AI projects in finance, using innovative machine learning models for stock price predication and auto-trading, outperforming conventional approaches.

Li will present two of his earlier AI finance projects. The first used a deep-learning long short-term memory model to forecast the alpha returns of approximately 1000 stocks on both the Chinese A-stock market and the U.S. stock market. Based on predicted returns, the research team devised a new portfolio strategy that outperformed a traditional model in both markets significantly.

In the second project, Li and his fellow researchers developed an optimal portfolio execution system (OPEX), using a tree-based ensemble machine-learning model for automated stock trading to reduce the trading cost. The study demonstrated that the OPEX system could effectively reduce trading costs, with an estimated savings of approximately $35 million per year compared to a legacy linear model.

Dr. Timothy (Minghai) Li currently serves as a principal data scientist and vice president at the central modeling team of Citizens Bank. Prior to this role, he held positions as a senior data scientist at Fidelity Investments, Netbrain Tech Inc., and FIS. He earned his Ph.D. in physics from Boston University and conducted post-doctoral research in Professor Sharon Huo’s group at Clark. He has authored or co-authored more than 20 publications covering topics in physics, material science, chemistry, biology, and computer modeling and simulation. Presently, his focus lies in the application of advanced machine learning algorithms and generative AI to finance.


April 3, 2024
1:30 p.m. - 2:30 p.m.
Event by Topic