For any level , based on the evaluated indices of previous level ( for the lowest one , based on the normalized data ) , combining the incomplete weight information of the level , a dea model with output only and preference information is was produced , and then mean cross - efficiencies were calculated 對于任意一層指標,以前一級的評價指數(shù)為基礎(chǔ),結(jié)合該層指標權(quán)重的不完全信息,建立只有輸出指標且含有偏好信息的數(shù)據(jù)包絡(luò)分析( dea )模型,求出平均交叉效率,并將其作為該層的評價指數(shù)。
Through detailed analysis to information entropy concept of complete information systems , weighted information entropy and weighted conditional entropy of incomplete information system based on limited tolerance relation are studied , which is naturally genealization of rough entropy conept of complete information systems based on equivalence relations 在仔細分析完備信息系統(tǒng)的信息熵概念的本質(zhì)之后,通過引入權(quán)數(shù),研究了基于限制容差關(guān)系的不完備信息系統(tǒng)加權(quán)信息熵和條件熵問題。