Article ID: | iaor200461 |
Country: | United Kingdom |
Volume: | 13 |
Issue: | 2 |
Start Page Number: | 68 |
End Page Number: | 76 |
Publication Date: | Mar 2000 |
Journal: | Logistics Information Management |
Authors: | Raggad Bel G. |
Keywords: | artificial intelligence: decision support |
Proposes a possibilistic group support system (PGSS) for the retailer pricing and inventory problem when possibilistic fluctuations of product parameters are controlled by a set of possibilistic optimality conditions. Experts in various functional areas convey their subjective judgement to the PGSS in the form of analytical models (for product parameters estimation), fuzzy concepts (facts), and possibilistic propositions (for validation and choice procedures). Basic probability assignments are used to elicit experts' opinions. They are then transformed into compatibility functions for fuzzy concepts using the falling shadow technique. Evidence is processed in the form of fuzzy concepts, then is rewritten back to basic probability assignments using the principle of least ignorance on randomness. The PGSS allows the user (inventory control) to examine a trade-off between the belief value of a greater profit and a lower amount of randomness associated with it.