A learning system for communications network configuration

✍ Scribed by Andrew Jennings

• Publisher: Elsevier Science
• Year: 1988
• Tongue: English
• Weight: 880 KB
• Volume: 1
• Category: Article
• ISSN: 0952-1976
• DOI: 10.1016/0952-1976(88)90001-2

✦ Synopsis

One of the major goals of artificial intelligence research is to create learning systems that are of practical use. To this end a great deal of effort has been expended in understanding and constructing learning mechanisms. This paper describes a learning system that is of practical use in the domain of communications network design and configuration. In constructing this system we have had to confront many of the issues for learning systems in poorly understood domains. Some important lessons for learning systems are presented. It is argued that in poorly understood domains a casebased learning approach is useful when implemented with careful regard for computational costs.