在「估測環境」下之模糊多準則決策模式 ; Fuzzy Multiple Attribute Decision Making Model under Stochastic Environment
Saved in:
| Title: | 在「估測環境」下之模糊多準則決策模式 ; Fuzzy Multiple Attribute Decision Making Model under Stochastic Environment |
|---|---|
| Authors: | 朱大中, Chu, Ta-Chung |
| Contributors: | 工業工程與管理系 |
| Publication Year: | 1999 |
| Collection: | Southern Taiwan University Institutional Repository (STUTIR) |
| Subject Terms: | 模糊多屬性決策, 隨機環境, 相容性函數, Fuzzy multiple attribute decision making, Stochastic environment, Compatibility function |
| Time: | 38 |
| Description: | 「模糊多準則決策」是學術界和企業界在研究問題和解決問題上所採用的重要決策模式之一。在「廠址評選」、「R&D計畫評選」、「產品品質評估」、「股票選擇」.等各方面均有廣泛的研究和應用。目前,在模式中分析「方案」(Alternatives)時,一般先討論出數個應評估之主因素(或準則),再分別由各主因素中篩選出數個次因素;其次,將主因素及次因素予以模糊化並輔以適當加權運算後,再做優先順序(Ranking)之處理以制定決策。本研究計畫則在兩方面考慮將上述模式予以擴充。第一,在因素方面:數量化因素需應用適當之Compatibility functions 予以模糊化,非數量化因素通常以TFN's作模糊化。然而,非數量化因素常因缺乏客觀評估標準而致「語意變數值」之給予非常主觀。本研究提出在一組語意變數值下,決策者在「估測環境」中對因素作評估,以解決在現實世界中,決策者可能給予因素作出不同語意評估之情況。例如評估汽車品質時,其中一項非數量化因素為「外型」(Body type),決策者有70%possibility評定為「好」及30%possibility評定為「普通」。第二,在權數方面:權數之給予亦因失之主觀而以TFN's處理,但決策者仍會因所設定的「語意變數值」不敷使用,或因個人偏好,或對因素認識不清等,而給予因素不同之語意評估。因此,本研究考慮在「估測」及「模糊」雙重條件下來建構權數模式,其特性是求層級架構中因素之語意權數之期望值。經上述第一及第二方面之考慮後,一嶄新的模糊多準則決策模式可以產生。最後,進行實例測試,並以Visual Basic 5.0 for Windows 95執行程式化。 ; Fuzzy Multiple Attribute Decision Making (FMADM) is one of the important decision making models on research and problem solving for the academic world and the enterprise circles. The applications of FMADM are very comprehensive, for example, "the location choice of a production", "the evaluation and choice on R&D problems", "product quality evaluation", "stock selection",. etc. So far, several needed major factors(or criteria) and their subfactors are usually selected firstly for fuzzification when analyzing the alternatives in the FMADM model. And then, weights are added in the model for operation. Finally, the ranking method is applied to solve fuzzy numbers (TFN's) for decision making. This research proposal is to expand the FMADM model by two ways. The first way is about the factors in the hierarchy structure. The quantitative factors need to be fuzzified by proper compatibility functions and the non-quantitative factors are usually fuzzified by TFN's. However, it is very subjective to give linguistic variable values due to lack of objective evaluation standard for non-quantitative factors. This research proposes that the decision maker evaluates factors by defined linguistic variable values in stochastic environment to solve the situation of different evaluation made by decision maker for factors in the real ... |
| Document Type: | other/unknown material |
| File Description: | application/pdf; 634856 bytes |
| Language: | Chinese English |
| Relation: | http://ir.lib.stust.edu.tw/handle/987654321/2566 |
| Availability: | http://ir.lib.stust.edu.tw/handle/987654321/2566 |
| Accession Number: | edsbas.F5ADBCE |
| Database: | BASE |
Nájsť tento článok vo Web of Science | Abstract: | 「模糊多準則決策」是學術界和企業界在研究問題和解決問題上所採用的重要決策模式之一。在「廠址評選」、「R&D計畫評選」、「產品品質評估」、「股票選擇」.等各方面均有廣泛的研究和應用。目前,在模式中分析「方案」(Alternatives)時,一般先討論出數個應評估之主因素(或準則),再分別由各主因素中篩選出數個次因素;其次,將主因素及次因素予以模糊化並輔以適當加權運算後,再做優先順序(Ranking)之處理以制定決策。本研究計畫則在兩方面考慮將上述模式予以擴充。第一,在因素方面:數量化因素需應用適當之Compatibility functions 予以模糊化,非數量化因素通常以TFN's作模糊化。然而,非數量化因素常因缺乏客觀評估標準而致「語意變數值」之給予非常主觀。本研究提出在一組語意變數值下,決策者在「估測環境」中對因素作評估,以解決在現實世界中,決策者可能給予因素作出不同語意評估之情況。例如評估汽車品質時,其中一項非數量化因素為「外型」(Body type),決策者有70%possibility評定為「好」及30%possibility評定為「普通」。第二,在權數方面:權數之給予亦因失之主觀而以TFN's處理,但決策者仍會因所設定的「語意變數值」不敷使用,或因個人偏好,或對因素認識不清等,而給予因素不同之語意評估。因此,本研究考慮在「估測」及「模糊」雙重條件下來建構權數模式,其特性是求層級架構中因素之語意權數之期望值。經上述第一及第二方面之考慮後,一嶄新的模糊多準則決策模式可以產生。最後,進行實例測試,並以Visual Basic 5.0 for Windows 95執行程式化。 ; Fuzzy Multiple Attribute Decision Making (FMADM) is one of the important decision making models on research and problem solving for the academic world and the enterprise circles. The applications of FMADM are very comprehensive, for example, "the location choice of a production", "the evaluation and choice on R&D problems", "product quality evaluation", "stock selection",. etc. So far, several needed major factors(or criteria) and their subfactors are usually selected firstly for fuzzification when analyzing the alternatives in the FMADM model. And then, weights are added in the model for operation. Finally, the ranking method is applied to solve fuzzy numbers (TFN's) for decision making. This research proposal is to expand the FMADM model by two ways. The first way is about the factors in the hierarchy structure. The quantitative factors need to be fuzzified by proper compatibility functions and the non-quantitative factors are usually fuzzified by TFN's. However, it is very subjective to give linguistic variable values due to lack of objective evaluation standard for non-quantitative factors. This research proposes that the decision maker evaluates factors by defined linguistic variable values in stochastic environment to solve the situation of different evaluation made by decision maker for factors in the real ... |
|---|