| Automated Synthesis of Multi-Agent Control |
14 SEP 2000 |
31 pages |
| Authors:
Maja J. Mataric; UNIVERSITY OF SOUTHERN CALIFORNIA LOS ANGELES DEPT OF COMPUTER SCIENCE
|
 | We have been pursuing a synthetic approach to studying the problem of controlling complex multi-robot systems by simultaneously developing a theory and testing it on complex domains consisting physical mobile robots. This process allows us to evaluate, improve, and further develop our theory, while producing a set of useful software and hardware applications. Our approach is behavior-based; the robots use a set of behaviors (parametric, goal-achieving ... |
|
| Proceedings of the Fourth International Conference on Simulation of Adaptive Behavior: From Amimals to Animals 4 |
DEC 96 |
645 pages |
| Authors:
Pattie Maes; Maja J. Mataric; Jean-Arcady Merer; Jordan Pollack; Stewart W. Wilson; BRANDEIS UNIV WALTHAM MA
|
| Collection of papers refereed and presented at the 'SAB96' Conference. |
|
| Interaction and Intelligent Behavior |
AUG 94 |
191 pages |
| Authors:
Maja J. Mataric; MASSACHUSETTS INST OF TECH CAMBRIDGE ARTIFICIAL INTELLIGENCE LAB
|
| We introduce basic behaviors as primitives for control and learning in situated, embodied agents interacting in complex domains. We propose methods for selecting, formally specifying, algorithmically implementing, empirically evaluating, and combining behaviors from a basic set. We also introduce a general methodology for automatically constructing higher--level behaviors by learning to select from this set. Based on a formulation of reinforcement learning using conditions, behaviors, and shaped reinforcement, out approach makes ... |
|
| A Comparative Analysis of Reinforcement Learning Methods |
OCT 91 |
|
| Authors:
Maja J. Mataric; MASSACHUSETTS INST OF TECH CAMBRIDGE ARTIFICIAL INTELLIGENCE LAB
|
 | This paper analyzes the suitability of reinforcement learning for both programming and adapting situated agents. In the the first part of the paper we discuss two specific reinforcement learning algorithms: Q-learning and the Bucket Brigade. We introduce a special case of the Bucket Brigade, and analyze and compare its performance to Q-learning in a number of experiments. The second part of the paper discusses the key problems of reinforcement learning: ... |
|
| A Distributed Model for Mobile Robot Environment-Learning and Navigation |
MAY 90 |
|
| Authors:
Maja J. Mataric; MASSACHUSETTS INST OF TECH CAMBRIDGE ARTIFICIAL INTELLIGENCE LAB
|
| This thesis presents a method for robust mobile robot navigation, large space learning, and path planning, based on a totally distributed architecture. The described methods were implemented and tested on a physical robot. The robot, Toto, consists of an omnidirectional base supplied with a ring of twelve ultrasonic ranging sensors and a compass. It is fully autonomous with all power and processing onboard. All experimental data were gathered in unaltered ... |
|
Reports by Author
