A team from Harvey Mudd College was named a finalist in this year’s Mathematical Contest in Modeling (MCM), which ran from Feb. 18 to Feb. 22. Harvey Mudd entered six teams in this year’s contest, with participants ranging from freshmen to seniors.
The team comprised of Richard Bowen HM ’10, Brett Cooper HM ’10, and Bryce Lampe HM ’10 made it to the finals, placing them in the top 24 teams out of 2,254 from 14 different countries.
In previous years, the school has performed equally well or better, receiving the highest title of “Outstanding”—usually awarded to only one or two percent of teams—in three of the last five annual contests. Professor Susan Martonosi, one of the faculty advisors for the Mudd teams, said the college has the highest number of “Outstanding” designations in the contest’s history, even though the school is a comparatively small institution.
“The contest stresses what Mudd does best, tying the mathematics or science or engineering to the world around us,” she said.
The MCM is an international contest administered via the internet. It spans 96 hours, during which students choose one of two open-ended modeling questions from the website and then respond with a short technical paper and summary. Participants may use any resources they like, in an attempt to make the contest as realistic and relevant as possible.
“As campus coordinator, I solicited participants, enrolled them in the contest, arranged rooms and computing services for them with the help of our math department staff, and made sure the students submitted their papers on time and in the required format,” Martonosi said. “The students did the rest.”
The contest is run by the Consortium of Mathematics and Its Applications (COMAP). The Outstanding contestants are chosen by The Institute for Operations Research and the Management Sciences (INFORMS), the Society for Industrial and Applied Mathematics (SIAM), and the Mathematical Association of America (MAA).
Bowen, Cooper, and Lampe prepared extensively for the contest by developing their workroom. They wheeled in two computer towers and monitors, a set of speakers, and a mattress on a dolly. They also bought a slew of snacks and energy drinks.
“We evaluated the cost-effectiveness of energy drinks, and ended up replicating 5-Hour Energy for cheaper,” Cooper said.
By doing so, they decreased spending by 67 percent.
“Someone sent out the papers of the previous contest winners, but we didn’t read them,” he said.
They agreed that their general math, probability and statistics, and Fourier classes prepared them well. They added the experience they gained working on their clinic projects—a Harvey Mudd capstone project for computer science and engineering majors—was their most crucial asset.
“It helped us learn how to work on one problem for a long period of time with a team,” Cooper said.
They were also glad they had an understanding of Bayesian probability, the programming language Python, and diverse coding backgrounds in general.
“We did a good job of starting on writing early, but we still had to make changes at the last minute,” Lampe said.
One thing they forgot to change was the temporary title they used as a placeholder, which read “Science of the Lambs.”
The problem chosen by Mudd’s team involved pinpointing the most likely position of a serial killer, given the locations of his previous killings.
“First, we got data, then we looked at every possible explanation we could come up with of the data,” Bowen said. “The hard part is justifying and explaining probabilities, and making it all clear.”
They started by searching for literature and previous research on their topic, which took approximately 12 hours. They then brainstormed model ideas, combining models they found as well as formulating their own.
“There was a lot of whiteboard work, and a few nasty integrals we didn’t know how to solve,” Bowen said.
The next step, they said, consisted of coding, debugging, writing and thinking of ways to justify their methods and results. They found themselves with a paper they considered mathematically sound, along with a great deal of justification.
“I never wanted to look at that problem again,” Bowen said of the end of the contest.