博碩士論文 104423051 完整後設資料紀錄

DC 欄位 語言
DC.contributor資訊管理學系zh_TW
DC.creator陳竑維zh_TW
DC.creatorHong-Wei Chenen_US
dc.date.accessioned2017-8-2T07:39:07Z
dc.date.available2017-8-2T07:39:07Z
dc.date.issued2017
dc.identifier.urihttp://ir.lib.ncu.edu.tw:88/thesis/view_etd.asp?URN=104423051
dc.contributor.department資訊管理學系zh_TW
DC.description國立中央大學zh_TW
DC.descriptionNational Central Universityen_US
dc.description.abstract決策是從所有的可行方案中找出最佳選擇的過程。不管簡單或複雜,我們的生活中都需要面對各式各樣的決策。決策的範疇可以從一早起床決定今天穿什麼樣的衣服這種簡單的決策到公司管理者制定可能會影響公司未來發展的決策,而隨著決策層級越高越複雜時,代表著所被牽涉到的相關影響因素也越多,決策的制定也就更加複雜與困難。三位學者Chen.、Huang.與Chang.在2015年提出了提出透過摘要化技術去幫助處理多準則決策的問題,利用摘要化技術產生決策彙整表,提供決策者相關的決策輔助資訊。而該研究中所設計的演算法隱含著一個問題:使用者必須事先定義替代方案與決策準則各別的集群值,而這個兩個數值往往是難以決定的,也同時影響著最後所產生的決策彙整表之表現。 本研究設計一個啟發式演算法,該演算法在歸納決策彙整表時,可以自動決定集群的數目,並使歸納的過程中所導致的資訊遺失量越小越好,也代表著最後的結果可以更真實貼近原始決策資訊讓決策者能夠依照自身的經驗及決策彙整表的輔助以制定更精確的決策。接著更近一步設計出一個Incremental方法應用於所提出的啟發式演算法中進行計算效率改良,最後透過基因演算法中的輪盤法來獲取更佳的結果。在實驗設計中透過九種決策表組合,其中每種決策表組合各包含十個資料集,總共九十個資料集來驗證本研究所提出的方法,並且導入了實際案例來闡述實驗的結果。zh_TW
dc.description.abstract Decision-making is a process to find the best option when there are many available alternatives. Decision happens in our daily life, no matter how simple or difficult it is. The coverage of decision-making can be from a simple decision such as what to eat for breakfast to complicated business decisions affecting the future development of companies. In general, the bigger the impact is, the more related factors are involved in a decision. This makes decision-making become difficult and complicated. Recently, Chen, Huang, and Chang. (2015) did a great work “Using Summarization Techniques to Resolve the Multi-Criteria Decision Making Problems” that using summarization techniques solves MCDM problems. However, there is a weakness in their proposed algorithm, i.e., users need to predefine the numbers of alternative and criteria clusters before executing their algorithm to summarize information. Unfortunately, the numbers are difficult to decide in practice and it can affect the performance of summarization table in the end. In this research, we first propose one heuristic algorithm to decide the numbers of clusters when summarizing decision tables. Next, we design a specific incremental method and apply it to the heuristic algorithm to improve the calculation efficiency. Finally, Genetic Algorithm – Roulette Wheel Selection is included into the algorithm to obtain better results. In experimental design and results, to verify the proposed approach, nine combinations of decision tables are used, each of them consisted of ten data sets. The results indicated that our algorithms can effectively and efficiently determine the numbers of alternative and criteria clusters when summarize the decision table. Moreover, a real case study is used to illustrate the effectiveness of the proposed approach.en_US
DC.subject多準則決策彙整表zh_TW
DC.subject決策問題zh_TW
DC.subject多目標決策zh_TW
DC.subject基因演算法zh_TW
DC.subject輪盤法zh_TW
DC.subjectSummarizing Multi-Criteria Decision Tablesen_US
DC.subjectDecision Making Problemen_US
DC.subjectMulti-Criteria Decision Makingen_US
DC.subjectGenetic Algorithmsen_US
DC.subjectRoulette Wheel Selectionen_US
DC.titleImproving Algorithms for Summarizing Multi-Criteria Decision Tablesen_US
dc.language.isoen_USen_US
DC.type博碩士論文zh_TW
DC.typethesisen_US
DC.publisherNational Central Universityen_US

若有論文相關問題,請聯絡國立中央大學圖書館推廣服務組 TEL:(03)422-7151轉57407,或E-mail聯絡  - 隱私權政策聲明