Poker Academy, a software developer of poker products, has donated
$13,000 in software to Lehigh University for students in computer
science and engineering to use in artificial intelligence (AI) research
projects.
Hector Munoz-Avila, assistant professor of computer science and engineering
at the University, said the donated software will be used in "Artificial
Intelligence: Game Programming," a course offered last
fall for the first time that will be taught again next fall.
The goal of the students using the donated software in the new class,
which Munoz-Avila teaches, will be to make better, more challenging
computer games that adapt to a player's behavior, style and level
of skill.
"Only a few games currently employ adaptive artificial
intelligence, which is also called machine learning,"
said Munoz-Avila, who has a grant from the Naval Research Laboratory
to study game programming. "Most are 'hard-coded' to operate
at a fixed level or levels.
"Adaptive AI is code that allows computer software to adapt
over time to a players' skill levels and even to improvements they
make," said Munoz-Avila. "A chess game equipped
with adaptive AI will 'dumb down' in a few moves against an inexperienced
player, but play tougher against a grandmaster."
"We are pleased to be part of this on-going research and
development initiative of artificial gaming intelligence,"
said Kurt Lange, president of Poker Academy. "The AI gaming
platform for our poker product, which was developed over more than
10 years, has computer opponents that actually think like humans.
Since the game gets better as the player improves, the game is always
challenging and thus a very effective learning tool. We look forward
to seeing what the CSE students come up with."
The students in Munoz-Avila's game-programming class study and develop
algorithms, which are methods of encoding programs that tell computers
how to solve problems. The students will use their algorithms to study
the software robots, or bots, that are created to replace human beings
in poker, hearts and other online games.
Other goals of the students will be to determine how "smart"
computers can be made and whether AI gives unfair advantages to the
bots or machines against which humans play computer games.
Munoz-Avila said that Jarret Raim, a graduate student in computer
science, found the Poker Academy online, looked at its API (application
program interface, a way in which programs communicate), and found
it could "talk" to other programs, making it a good fit
for the Lehigh game programming course.
"Our interest is that this game has a very nice API that
we can interface with our machine learning programs," said
Munoz-Avila.