查詢結果分析
來源資料
相關文獻
- 運用約略集合理論建構特殊軍品管理知識庫
- A Constraint-Based Genetic Algorithm for Data Mining
- The Genetic Algorithm with Rough Set Theory Incorporated into the Patent Composition
- 應用約略集合理論建立電磁干擾診斷系統
- 應用約略集合理論於醫療機構商業流程之委外供應商評選
- Product Bundling in the Electronic Commerce Environment: A Hybrid Approach
- A Hybrid Approach for Knowledge Recommendation
- 運用資料探勘輔助商品分類之需求預測方法
- 模糊約略集合法於顧客與購買樣式之關聯分析
- 網路算命使用者行為與特徵分析:資料探勘技術之應用
頁籤選單縮合
題名 | 運用約略集合理論建構特殊軍品管理知識庫=Applying Rough Set Theory to Constructing Knowledge Base for Critical Military Commodity Management |
---|---|
作者 | 邱華凱; 黃雍婷; Chiou, Hua-kai; Huang, Yong-ting; |
期刊 | 中華科技大學學報 |
出版日期 | 20091100 |
卷期 | 41 2009.11[民98.11] |
頁次 | 頁171-186 |
分類號 | 595 |
語文 | chi |
關鍵詞 | 約略集合理論; 資料探勘; 德爾菲法; 屬性折減; 基因演算法; 特殊軍品管理; Rough set theory; Data mining; Delphi process; Attribute reducts; Genetic algorithm; Critical military commodity management; |
中文摘要 | 約略集合理論(Rough Set Theory)係基於規則的資料探勘與人工智慧的演算 法,適合用來尋找不精確、不完整資料所隱含的知識,其特色為不需對所屬資料 集有先驗或是額外的資訊,且資料呈現不需符合任何統計假設。許多特殊軍品諸 如彈藥與炸藥,其儲存地區存在一定程度的危險性,倘若彈藥與炸藥的儲備品不 小心被遺棄、毀損,或者不適當地儲存與管理,對整個儲存區域將具有相當程度 的威脅。隨著時代進步,國內人口與建築物的密集度愈來愈高,故當意外事故肇 生後,其影響所及範圍的傷亡程度可能較以往更為嚴重。本研究選定國軍彈藥安 全管理為主題,首先運用德爾菲法將評估委員之意見歸納出影響彈藥安全管理之 6項主要構面包括39項關鍵因素,再以約略集合理論結合基因演算法與分解樹分 類法找出相關屬性之重要性權重,並以約略集合理論進行特殊軍品管理決策之分 類,透過其屬性折減的知識學習特性,成功將1187筆彈藥管理決策資料,分別以 「處理方式」與「危險等級識別」等兩項核心決策變數,萃取出351條與362條精 簡的決策法則,並歸納出六項共同的核心屬性,以提高管理的安全性與便利性。 本研究發現,採取基因演算法進行屬性折減所找出的關鍵要素,可提高對特殊軍 品管理分類的準確性,作為相關單位於危險品管理與決策之參考。 |
英文摘要 | Rough Set Theory prospects the algorithm of performing with data mining and artificial intelligence. It is suitably applied to mining the implicit knowledge to dealing with imprecise or incomplete data. Its characteristic, for not needing to collect priori or extra information to the affiliated materials, and the materials do not need to accord with any statistical hypothesis. Such as the ammunition and explosive of a lot of special military products, the danger of a certain degree that it stores the area and exists, if the reserve products of the ammunition and explosive are abandoned, damaged because of carelessness, or store and manage inappropriately, to store area have a threat analogous to the degree while being whole. The progress with era, domestic population and intensive degree of the building are higher and higher, so after the contingency is started and grown, the injuries and deaths degree of the range may be more serious than in the past that it is influenced in one's power. This research selects army's ammunition safety management definitely as the empirical case; we present applications of rough set theory for establishing the knowledge base by pattern recognition. Firstly, we employ Delphi process to converge 39 critical attributes with respect to 6 key factors for feature selection. We then utilize the rough set approach combing with genetic algorithm and decomposition tree method extracting 351 and 362 decision rules simplified with such two key decision parameters, treatment and classification of hazardous substance, and 6 core reducts induced from 1187 ammunition administrative decision materials, to improve the security and convince for ammunition safety management We further stated some concluding remarks in the final section. Through this research we found that rough set theory is an efficiency technique for pattern recognition, it can improve and manage the accuracy classified to the critical military commodity, in hazardous articles management and reference of decision as the relevant authority department. |
本系統之摘要資訊系依該期刊論文摘要之資訊為主。