U of T students Sean Yi, Serena Ban, and Marvin Guo win Undergraduate Award at Forecasting Competition

A team of talented undergraduate students from the University of Toronto has won the Undergraduate Award at the 2023 CANSSI Ontario French Trot Horse Racing Forecasting Competition. The award recognizes their great work in developing a forecasting model, earning them a $1,500 prize.*

The team, consisting of Sean Yi (Department of Statistical Sciences (DoSS) and the Department of Computer Science), Serena Ban (Departments of Chemistry DoSS), and Marvin Guo (DoSS), demonstrated their expertise in data analysis, machine learning, and innovation. The competition challenged participants to create effective models for predicting outcomes in French Trot Horse Racing, emphasizing both performance and innovation.

Innovative Approach to Predictive Modeling. To tackle the complex dataset provided, the BGY Team applied a rigorous methodology:

Data Cleaning and Feature Selection: The team analyzed the dataset, identified and removed duplicate or irrelevant features, and refined the data to ensure its quality.
Feature Engineering: Using logistic regression, the team focused on eight key performance-related features, such as “Prizemoney” and “FinishPosition,” to develop a new feature, “WinningProbability.” This variable became the predictive target for their model.

Advanced Modelling: To optimize accuracy, they employed a Stacking Regressor that combined a Multilayer Perceptron (MLP) Regressor and a Random Forest Regressor. This innovative approach leveraged the strengths of both models to produce superior forecasting results.

About the Competition

The French Trot Horse Racing Forecasting Competition, launched by CANSSI Ontario in 2023, invited undergraduates, graduate students and postdoctoral fellows across Ontario to showcase their skills in statistics and machine learning. Participants were tasked to build the best models for accurately predicting French Trot Horse Racing. The competition evaluated submissions across three categories:

• Performance: Achieving the most accurate winner forecasts on a three-month validation dataset ($5,000).
• Innovation: Developing novel features, exploring new methodologies, or presenting compelling narratives ($3,500).
• Best Entry by Undergraduates: Undergraduates were judged in either category—performance or innovation—the award was reserved for the best undergraduate entry ($1,500).

“We are thrilled to have had the opportunity to participate in such a unique competition,” said team member Sean Yi. “This experience allowed us to apply our knowledge of statistics and machine learning to a real-world problem, and we’re incredibly proud of what we accomplished as a team.”

For more information about the 2023 CANSSI Ontario Forecasting Competition, visit the Competition webpage.

About the University of Toronto

The University of Toronto is one of the world’s leading institutions in research and education, renowned for its innovation and excellence in a wide range of fields, including statistical sciences and computer science.

About CANSSI Ontario

The Canadian Statistical Sciences Institute (CANSSI) Ontario fosters collaboration in the statistical sciences across Canada, supporting research and professional development in the field.