TY - GEN
T1 - Learning to win games in a few examples
T2 - 10th International Conference on Social Robotics, ICSR 2018
AU - Ayub, Ali
AU - Wagner, Alan R.
N1 - Funding Information:
Acknowledgement. Support for this research was provided by Penn State’s Teaching and Learning with Technology (TLT) Fellowship.
Publisher Copyright:
© 2018, Springer Nature Switzerland AG.
PY - 2018
Y1 - 2018
N2 - Teaching robots new skills using minimal time and effort has long been a goal of artificial intelligence. This paper investigates the use of game theoretic representations to represent interactive games and learn their win conditions by interacting with a person. Game theory provides the formal underpinnings needed to represent the structure of a game including the goal conditions. Learning by demonstration, has long sought to leverage a robot’s interactions with a person to foster learning. This paper combines these two approaches allowing a robot to learn a game-theoretic representation by demonstration. This paper demonstrates how a robot can be taught the win conditions for the game Connect Four using a single demonstration and a few trial examples with a question and answer session led by the robot. Our results demonstrate that the robot can learn any win condition for the standard rules of the Connect Four game, after demonstration by a human, irrespective of the color or size of the board and the chips. Moreover, if the human demonstrates a variation of the win conditions, we show that the robot can learn the respective changed win condition.
AB - Teaching robots new skills using minimal time and effort has long been a goal of artificial intelligence. This paper investigates the use of game theoretic representations to represent interactive games and learn their win conditions by interacting with a person. Game theory provides the formal underpinnings needed to represent the structure of a game including the goal conditions. Learning by demonstration, has long sought to leverage a robot’s interactions with a person to foster learning. This paper combines these two approaches allowing a robot to learn a game-theoretic representation by demonstration. This paper demonstrates how a robot can be taught the win conditions for the game Connect Four using a single demonstration and a few trial examples with a question and answer session led by the robot. Our results demonstrate that the robot can learn any win condition for the standard rules of the Connect Four game, after demonstration by a human, irrespective of the color or size of the board and the chips. Moreover, if the human demonstrates a variation of the win conditions, we show that the robot can learn the respective changed win condition.
UR - http://www.scopus.com/inward/record.url?scp=85058348297&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85058348297&partnerID=8YFLogxK
U2 - 10.1007/978-3-030-05204-1_34
DO - 10.1007/978-3-030-05204-1_34
M3 - Conference contribution
AN - SCOPUS:85058348297
SN - 9783030052034
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 349
EP - 358
BT - Social Robotics - 10th International Conference, ICSR 2018, Proceedings
A2 - Broadbent, Elizabeth
A2 - Ge, Shuzhi Sam
A2 - Salichs, Miguel A.
A2 - Castro-González, Álvaro
A2 - He, Hongsheng
A2 - Cabibihan, John-John
A2 - Wagner, Alan R.
PB - Springer Verlag
Y2 - 28 November 2018 through 30 November 2018
ER -