Although Anna Eilering and Nick Becker came to OSU's workshop on Monte-Carlo methods with vastly different experiences in artificial intelligence, they said they both learned a lot by working together to create an artificial "agent" to play the game Galcon.
The rapidly growing field of artificial intelligence (AI) has broad applications in computer games, robotics, air traffic control, ecological modeling and many other areas. To inspire more interest for AI, Oregon State University and the National Science Foundation have teamed up for two years to hold an all-expense-paid workshop for undergraduate students from around the country.
“Artificial intelligence and machine learning are really hot these days and they are applied in most every area you can think of. It’s very important for the ideas to come out of the laboratories and into the market place, and in order to be able to do that we need to have a lot more students,” said Prasad Tadepalli, professor of computer science and co-organizer of the course.
The School of Electrical Engineering and Computer Science hosted the program to encourage more students to seek advanced degrees in computer science, especially women and minorities who are under-represented in the field.
This spring, twenty-one students had the opportunity to learn from top AI researchers and experiment with state-of-the-art research tools. Professors gave the students a foundation on Monte Carlo methods in artificial intelligence and exposed them to current areas of research.
Olga Torres was inspired by the competition to learn more about artificial intelligence methods.
During the lab sessions, each student designed an artificial “agent” to play a real-time strategy game called Galcon. For the final project students worked in teams of two or on their own to create their best agent to battle it out in a tournament.
“By building on top of a home-grown software framework, the students had an opportunity to show their creativity in design and use the methods we’ve taught them to build more sophisticated agents,” Tadepalli said.
Many students commented that the competition spurred them on.
Olga Torres who had never programmed a game before said, “I really loved programming this game. I'm planning on continuing to work on it just for myself, to see if I can beat all the other agents. I'm very competitive and I just want to take over!”
Torres is interested in many branches of computer science which she currently studies at the College of William and Mary in Williamsburg, Va. The workshop gave her a deeper understanding of AI which will help her decide on a direction for her future.
“I probably wouldn't be able to get this much useful information in such a short period of time from anywhere else,” Torres said.
Beyond the game
Colin Gavin had a specific purpose for coming to the workshop. At Lewis and Clark in Portland, Ore., where he is a student of math and physics, he has been working on an artificial intelligence project to employ Monte Carlo methods to the board game “Go.”
Colin Gavin's previous experience with AI helped him win the Galcon tournament. He said the workshop was "very useful" and he learned about how Monte Carlo methods are used in applied research.
“I really wanted to get a firmer background on the Monte Carlo method and the workshop was very useful for that,” Gavin said.
But he also learned that the Monte Carlo methods are not just for games. “That's the setting that I had seen them in before, but we saw applications in robotics, ecology, civil works planning, and much more. It was really enlightening how broad and applicable these methods are,” Gavin said.
Anna Eilering, who is studying computer science with an emphasis on AI at Indiana University in Bloomington was excited to hear directly from AI researchers about how they are using these techniques.
“I'm trying to figure out if I want to do this for graduate school. So getting more experience in actual applications of artificial intelligence was really helpful,” Eilering said.
Eilering also enjoyed the experience of working with the other students.
“It was awesome how diverse the whole crowd was in that everybody had different academic backgrounds. There were linguistics majors, math majors, and people in cognitive science. It was really cool to see how everybody brought something different to the conversation,” Eilering said.
And just as Tadepalli had hoped, the experience inspired Eilering to consider Oregon State for her graduate work.
“I think it's an amazing school, with amazing professors. I'd really like to move out west and tinker with robots,” she said.