From predicting professional hockey outcomes to printing 3D models of crystal structures, undergraduate Brock students showcased their work at the recent Faculty of Mathematics and Science Undergraduate Research Symposium.
Nearly two dozen students exhibited posters outlining their research and discussed their projects with faculty members and peers during the event on Thursday, Aug. 10.
Kristen Rose, a second-year Mathematics and Statistics student and research assistant for Associate Professor William Marshall, shared her work in which she used regularized linear regression to predict professional hockey game outcomes and, in particular, how many goals top-scoring players will score in their next game.
Rose’s model considered statistics from the 2022-23 National Hockey League season and found that the top five factors for a player scoring goals are average time on ice, average goals, average power play time on ice, previous game plus/minus and the opponent’s average penalty minutes.
A key factor for players not scoring a goal was high average hits against their opponents.
“Hits are generally regarded as positive, but it actually influences not scoring goals,” said Rose. “Some possible reasons this might be, first off, if you go for a hit, you can’t go for a puck. Your teammates can go for a goal, which is why hits are generally considered a good thing, but you can’t score.”
Hits also lead to penalty minutes, which means less time on the ice.
“A player’s average time on ice is the highest factor for scoring goals, so if they have less time on the ice — if they get a penalty — they are less likely to score a goal,” she said.
Prediction models like Rose’s could be used to help with team strategy, determining players’ salaries, developing and training players, marketing a team around a top-scoring player and scouting new players.
For their research, first-year Physics student and research assistant Sebastien Duguay worked with Assistant Professor Gavin Hester to create 3D printed models of crystal structures to use in education and outreach.
“We wanted to find a way to make physical handheld models of the crystal structures we work with,” said Duguay. “In particular, we were interested in the Bravais lattices, which are the 14 possible ways atoms can arrange themselves in the three-dimensional crystal space.”
The intention for the 3D models is to help students and others learn about crystal structures, important foundational knowledge in a variety of sciences, including physics, chemistry and earth sciences.
“Crystal structures can be hard to visualize, especially the more complex ones,” Duguay said. “When they’re printed in a 3D model, you can much more easily see where the atoms are arranged on each of the axis.”
The 3D print files are available as open source on printables.com so other scientists and educators can access them.
All students who participated in the symposium were judged based on their presentation skills and their poster design. Winners received a certificate of accomplishment and a cash prize funded by the Faculty of Mathematics and Science Dean’s Office and Scion Instruments Canada.
Outstanding Poster Presentation Awards
- First place: Rhea Alitawi, Neuroscience — “Developmental and Cognitive Outcomes of Ethanol Exposure during Invertebrate Embryogenesis”
- Second place: Aws Al Jumaily, Computer Science — “Transformer-Based Multi-Objective Reinforcement Learning for Drug Design”
- Third place: Eric He, Physics — “Bound States without Potentials: Localization at Singularities”
Outstanding Poster Design
- George White, Earth Sciences — “Identifying Possible Wrinkle Ridge Exposures in the Walls of Valles Marineris, Mars”
- Lisa Harris, Chemistry — “Utilizing Historical Data to Predict Water Quality Trends of the Wignell Drain in Port Colborne”