Please use this identifier to cite or link to this item: https://hdl.handle.net/10356/17134
Title: Multiple-criterion decision making techniques for partner selection
Authors: Lau, Wei Wah.
Keywords: DRNTU::Engineering::Systems engineering
Issue Date: 2009
Abstract: In recent years, the infrastructure and logistical support for industries has improved tremendously. This has resulted in a competitive environment where suppliers can come from anywhere internationally. The Cognitive Advisor (CA) is a developed prototype which assists in selecting suitable suppliers based on contextual information and pre-determined business rules. It implements a hybrid of Analytic Hierarchy Process (AHP) and Deviation Measure techniques to rank candidate suppliers. CA is currently in its infancy status and requires a lot of beta testing. Furthermore, the results produced by the algorithms and business rules need to be validated. Therefore, in this project, the objective is to rigorously test the correctness and effectiveness of CA and provide an objective assessment of the way a suppliers’ suitability ranking is determined. To verify the correct execution of Cognitive Advisor, sample tests were conducted. The result of CA is compared to the scores obtained by Excel spreadsheet calculation in Microsoft Office Excel 2007. To investigate the effectiveness of CA, implementation of the ranking methods were evaluated and verified 2 test scenarios. This report provides a comprehensive description on the methodology of conducting the tests and detailed analysis on the results obtained. According to the analysis, irrelevant methods are identified. In test scenario 1, the slightly increase of payment duration from run 1 to run 8 has no significant impact on overall rankings due to low weight. In the real world scenario, capacity and payment penalty have more impact than other criteria as the application of deviation method will return highly positive (capacity) or highly negative (payment penalty) individual criterion cost. It causes rank reversal when 1 criterion performs superior and the rest performs poor and vice versa. Hence, alternative method should be researched in the future.
URI: http://hdl.handle.net/10356/17134
Rights: Nanyang Technological University
Fulltext Permission: restricted
Fulltext Availability: With Fulltext
Appears in Collections:MAE Student Reports (FYP/IA/PA/PI)

Files in This Item:
File Description SizeFormat 
mA211.pdf
  Restricted Access
3.33 MBAdobe PDFView/Open

Page view(s)

243
checked on Sep 28, 2020

Download(s)

18
checked on Sep 28, 2020

Google ScholarTM

Check

Items in DR-NTU are protected by copyright, with all rights reserved, unless otherwise indicated.