Learning automata and its application to priority assignment in a queuing system with unknown characteristics /
Abstract
Conditions for (epsilon)-optimality of a general class of absorbing barrier and strongly absolutely expedient learning algorithms are derived. As a consequence, a new class of learning algorithms having identical behavior under the occurrence of success and failure are obtained. An application of learning automata to the priority assignment in a queuing system with unknown characteristics is given.
Collections
- OU - Dissertations [9317]