A Study of Combining the Fuzzy linguistic preference relation and Artificial Neural Network for Multi-attribute Decision Making

碩士 === 國立高雄應用科技大學 === 企業管理系 === 98 === The analytical hierarch process (AHP) is commonly method used to making decisions. It can consider qualitative and quantitative factors in the same time. However, AHP still has few shortages; it can’t resolve ambiguity of problems, and not easy to consistency b...

Full description

Bibliographic Details
Main Authors: Mei-Chieh Ho, 何枚潔
Other Authors: Hui-Chung Yeh
Format: Others
Language:zh-TW
Published: 2010
Online Access:http://ndltd.ncl.edu.tw/handle/87300944337289897179
Description
Summary:碩士 === 國立高雄應用科技大學 === 企業管理系 === 98 === The analytical hierarch process (AHP) is commonly method used to making decisions. It can consider qualitative and quantitative factors in the same time. However, AHP still has few shortages; it can’t resolve ambiguity of problems, and not easy to consistency by increasing the number of criteria. In addition, most of decision problems need to predict the target value. But the decision-making method is rare which integrated two purposes. This study proposes a decision process which combines fuzzy LinPreRa with artificial neural network. Primarily, the expert opinions will be collected, and then use fuzzy LinPreRa to get the weights of fuzzy attributes. This method not only changes from fuzzy linguistic to quantitative, but also easy to using on select programs. If managers want to predict the estimated value of the programs, they can combine weights with artificial neural network to predict the target value. And the various programs will be more specifically in manager’s mind. Finally, this study tests the decision-making process by a supermarket chain in Kaohsiung. Primarily, this study uses fuzzy LinPreRa to get the weight to assessment of three locations. Afterward, this study use regression analysis and artificial neural network to predict turnovers. The results of the turnovers have same order. It showed this process can do location decision more effectively.