Publication View

Reinforcement Learning in the Sensor-Evader Domain (2009)

Abstract
Abstract. We present a study of reinforcement learning in the adversarial sensorevader domain, where a group of mobile sensors are trying to detect a moving evader. Amongst other results, we show that for the sensor, there are cases where following a learned policy does not significantly improve detector performance when compared to a random policy. 1

Publication details
Download http://citeseerx.ist.psu.edu/viewdoc/summary?doi=?doi=10.1.1.138.3016
Source http://www-users.cs.york.ac.uk/~kudenko/aamas05b.pdf
Contributors CiteSeerX
Repository CiteSeerX - Scientific Literature Digital Library and Search Engine (United States)
Type text
Language English
Relation 10.1.1.55.8066, 10.1.1.1.5959