Computer systems that learn: Sholom M. Weiss and Casimir A. Kulikowski

✍ Scribed by Alberto Segre; Geoffrey Gordon

• Book ID: 102990198
• Publisher: Elsevier Science
• Year: 1993
• Tongue: English
• Weight: 969 KB
• Volume: 62
• Category: Article
• ISSN: 0004-3702
• DOI: 10.1016/0004-3702(93)90084-o

✦ Synopsis

s book Computer ,Systems That Learn is a practitioner's guide to classification methods. Spanning methods from statistical pattern recognition, neural networks, and machine learning, its primary contribution is the way it draws these methods together in a uniform manner. By standardizing terminology and adopting a single evaluation criterion, Weiss and Kulikowski are able effectively to compare and contrast these different techniques. The book is a fine introduction to the area and could serve either as an introductory-level textbook or a handbook for analysts who wish to incorporate inductive learning techniques (e.g., classification methods) into their systems. Little or no previous background--save perhaps some elementary statistics--is assumed or required.

The book divides, roughly speaking, into three sections. In the first section (Chapters 1 and 2), the authors introduce the general classification model and outline an empirical methodology for estimating the true performance of a classification learning system. The second section (Chapters 3,4,and 5)