Files in this item

FilesDescriptionFormat

application/pdf

application/pdf9712259.pdf (8MB)Restricted to U of Illinois
(no description provided)PDF

Description

Title:Knowledge-guided constructive induction
Author(s):Donoho, Steven Kirk
Doctoral Committee Chair(s):Rendell, Larry A.
Department / Program:Computer Science
Discipline:Computer Science
Degree Granting Institution:University of Illinois at Urbana-Champaign
Degree:Ph.D.
Genre:Dissertation
Subject(s):Artificial Intelligence
Computer Science
Abstract:The relationship between constructive induction and domain knowledge can be analyzed systematically. Recent research has integrated knowledge and constructive induction in isolated domains with some success, but undue emphasis is often given to the technique developed rather than the underlying phenomena. In contrast, understanding how knowledge affects a domain's constructed feature space and instance space allows us to find better techniques more easily. Three case studies (bankruptcy, turfgrass management, and promoters) demonstrate how analysis of knowledge drives development of techniques. Although knowledge in each domain has its peculiarities, general relationships may be discovered that lead to a more principled approach to incorporating knowledge into constructive induction.
Issue Date:1996
Type:Text
Language:English
URI:http://hdl.handle.net/2142/22747
ISBN:9780591197990
Rights Information:Copyright 1996 Donoho, Steven Kirk
Date Available in IDEALS:2011-05-07
Identifier in Online Catalog:AAI9712259
OCLC Identifier:(UMI)AAI9712259


This item appears in the following Collection(s)

Item Statistics