Comparing an expert system vs traditional approach to forecasting item demand in a distribution inventory environment /
This research developed and tested an expert forecasting system using the C Language Integrated Production System (CLIPS) expert system shell. CLIPS allows the representation of expert knowledge as rules and allows the use of the object-oriented design paradigm. The testing of the expert system wa...
| Main Author: | |
|---|---|
| Format: | Thesis Book |
| Language: | English |
| Published: |
[Place of publication not identified] :
[publisher not identified] ;
1995.
|
| Subjects: | |
| Online Access: | http://proxy.library.tamu.edu/login?url=http://proquest.umi.com/pqdweb?did=742713141&sid=1&Fmt=2&clientId=2945&RQT=309&VName=PQD |
| Summary: | This research developed and tested an expert forecasting system using the C Language Integrated Production System (CLIPS) expert system shell. CLIPS allows the representation of expert knowledge as rules and allows the use of the object-oriented design paradigm. The testing of the expert system was undertaken under realistic business conditions using actual item demand data and an experienced human user. This research focused on the forecasting of inventory demand by a company providing a diverse range of industrial products to selling locations from a regional warehouse. A central issue explored by this research is whether or not the expert forecasting system can be used in a business environment to automatically generate forecasts for inventory items that will be at least as accurate as those generated by the forecasting system currently being used by the company participating in the test. In this study, none of the differences in forecasting accuracy among the various forecasting methodologies used were significant. However, the expert system's forecasting accuracy was better than that of the test company's in many comparisons. This may validate the expert system approach to forecasting. The expert system is constructed in such a way that the business user is able to specify that a particular forecasting method be used for a particular item or to alter the numerical forecast values generated by the expert forecasting system. In addition, the business user was allowed to select a forecasting method and then alter the numerical forecast values generated by that method. In this study, user input to the forecasting process did not create significantly better forecasting accuracy. However, in many cases in this study, the user's input resulted in the avoidance of large absolute forecasting errors and underforecasting. |
|---|---|
| Item Description: | Vita. "Major Subject: Business Analysis". |
| Physical Description: | xxi, 241 leaves : illustrations ; 28 cm. Issued also on microfiche from University Microfilms Inc. |
| Bibliography: | Includes bibliographical references. |