President's Showcase

Eli Butters

Poster Presentation, Ballroom D
Machine Learning Algorithms in Quantitative Finance
Supervising Professor: Dr. Adrian Barbu
Eli Butters is a junior majoring in Statistics and Computer Science and a member of the Florida State Varsity Swim & Dive team. His project involves discovering the best types of machine learning algorithms for different time series and classification problems. He was inspired to learn more about machine learning from recent breakthroughs in deep learning and their application in financial markets. Within his major, he has had the opportunity to learn more about how computers operate which has given him the knowledge to experiment with large deep learning models.


Machine learning can be applied in a variety of different ways within financial markets in order to capitalize on inefficiencies. This research explores 3 different models designed to learn the patterns of time series data and predict their movement in the coming time periods. These models have proved effective on basic datasets but fell short when given complex and noisy market data. Instead, problems are situated to form a binary classification issue rather than a regression problem. This solves the problem of the model fitting onto useless market noise. In order to do this, a pairs trading strategy originally published by Stanford researcher Jiayu Wu, is adapted. The original strategy uses an SVM to classify time periods as either converging or diverging using price as well as other technical indicators. During the development of the strategy many architectures and types of neural networks are tested in order to achieve more accurate results. This strategy allows for more accurate capitalization on these statistical arbitrage opportunities in the markets.

Presentation Materials

Project Materials

Project Documents and Links