Title :
Towards imitation-enhanced Reinforcement Learning in multi-agent systems
Author :
Erbas, Mehmet D. ; Winfield, Alan F T ; Bull, Larry
Author_Institution :
Bristol Robot. Lab., Univ. of the West of England, Bristol, UK
Abstract :
Imitation, in which an individual observes and copies another´s actions, is a powerful means of learning. This paper presents a way of using imitation to enhance the learning capability of individual agents. The agents employ Q-learning and we show that agents with imitation enhanced Q-learning learn faster than those with Q-learning alone.
Keywords :
learning (artificial intelligence); multi-agent systems; Q-learning; imitation-enhanced reinforcement learning; multi-agent systems; Actuators; Adaptation models; Electronic mail; Greedy algorithms; Learning; Robots; Watches;
Conference_Titel :
Artificial Life (ALIFE), 2011 IEEE Symposium on
Conference_Location :
Paris
Print_ISBN :
978-1-61284-062-8
DOI :
10.1109/ALIFE.2011.5954652
