Construction of an Imitation Learning Agent Using Game Records of Unspecified Players
Masayuki Ueno, Tomoyuki Takami · 2021 IEEE 10th Global Conference on Consumer Electronics (GCCE) · 2021
If it is possible to make a machine learn how a human learner behaves in a certain situation, it can be used as an agent that behaves like a human. Such an imitation learning agent can be used in various educational applications. We describe a learning method for constructing an imitation learning agent from a small number of game records of a specific individual, based on the game records of unspecified players.