We're interested in games, which are virtual environments that solve complex real problems instead.
The gaming environment is complex, but it helps us solve our problems quickly and easily.
We study the future.
We create artificial intelligence that looks at large rewards far away than small rewards in the environment. RL makes artificial intelligence that goes beyond people from ignorance.
We're interested in artificial neural networks.
Each neuron(node) makes it look like a human thought. Complex intertwining neurons produce better performance than previous algorithms.
We find value in data.
There's a lot of data that's not worth knowing until you look at it. We find great facts between them.
Research to solve complex video games. Solve complex problems by applying recognition models, in-depth structures, and multimodal to virtual environments based on Deep Reinforcement Learning.
Predict gamers' behavior based on multiple game logs. Preprocess and analyze the log data of the game company.
A project to conduct self-driving research on night/hazardous areas. Apply to the actual vehicle using models learned from reinforcement training in the simulator.
Interested in our research?
GIST, Republic of Korea
Room 519, Dasan Building
Prof. Kim Kyung-Joong