Journal article
AN EXPERIMENTAL INVESTIGATION OF THE PREDICTIVE ACCURACY OF INDUCTION AND REGRESSION
Expert systems with applications, v 7(4), pp 535-544
01 Oct 1994
Featured in Collection : UN Sustainable Development Goals @ Drexel
Abstract
As both research efforts and commercial adoption of Expert Systems (ESs) increase, knowledge acquisition continues to engage the attention of many ES developers. Many view the development of automated knowledge acquisition methods as one of the last remaining barriers to more widespread ES adoption and use. Rule induction has been advanced as a promising method of knowledge acquisition for expert systems, both to understand the weighting and importance of decision criteria and to provide a predictive capability for new examples. One important strand of research has naturally begun to compare the performance of rule induction to those of traditional statistical methods, such as multiple regression. In pursuing this subject, we document a comparison of the predictive capabilities of both induction and multiple regression. The study uses a sample of over 400 MBA students, employing their current GPA as the dependent variable or goal and prior (undergraduate) GPA and GMAT scores as independent or causal variables. An extensive comparison of both methods yields conclusions at variance with many previous studies which assert the superiority of the inductive method. Furthermore, the study identifies conditions under which the inductive method may be more effective as a predictor and develops recommendations for a contingency-based approach to knowledge acquisition for ES.
Metrics
Details
- Title
- AN EXPERIMENTAL INVESTIGATION OF THE PREDICTIVE ACCURACY OF INDUCTION AND REGRESSION
- Creators
- B Arinze - Drexel UniversityPNS Narasimha
- Publication Details
- Expert systems with applications, v 7(4), pp 535-544
- Publisher
- Elsevier
- Number of pages
- 10
- Resource Type
- Journal article
- Language
- English
- Academic Unit
- Decision Sciences (and Management Information Systems)
- Web of Science ID
- WOS:A1994PE16900006
- Scopus ID
- 2-s2.0-0039268229
- Other Identifier
- 991019318941304721
UN Sustainable Development Goals (SDGs)
This publication has contributed to the advancement of the following goals:
Source: SDGs in the Output
InCites Highlights
Data related to this publication, from InCites Benchmarking & Analytics tool:
- Web of Science research areas
- Computer Science, Artificial Intelligence
- Engineering, Electrical & Electronic
- Operations Research & Management Science