Communications of the ACM
Army Researchers Develop New Algorithms to Train Robots
Credit: Fortune
Researchers at the U.S. Army Research Laboratory (ARL) and the University of Texas at Austin have developed new methods for robots or computer programs to learn how to perform tasks by interacting with a human instructor.
The researchers focused on a case in which a human provides real-time feedback in the form of a critique. The Deep TAMER system employs deep learning to provide a robot with the ability to learn how to perform tasks by viewing video streams in a short amount of time with a human trainer.
The ARL/UT Austin team considered situations where a human teaches an agent how to behave by observing it and providing feedback, such as "good job," or "bad job." They demonstrated Deep TAMER's success by using it with 15 minutes of human-provided feedback to train an agent to perform better than humans on the Atari game of bowling.
From U.S. Army
View Full Article
Abstracts Copyright © 2018 Information Inc., Bethesda, Maryland, USA
No entries found