计算机技术与发展
計算機技術與髮展
계산궤기술여발전
COMPUTER TECHNOLOGY AND DEVELOPMENT
2014年
3期
39-41,45
,共4页
粗集%最简决策表%对象频度%约简%算法
粗集%最簡決策錶%對象頻度%約簡%算法
조집%최간결책표%대상빈도%약간%산법
rough set%simplest decision table%object frequency%reduction%algorithm
决策表中大量的不相容对象和相同对象会降低粗集约简算法性能，出现错误的属性核和约简结果。因此文中首先讨论决策表的不相容对象与相同对象即冗余对象对约简结果的影响，然后提出利用对象频度概念来删除冗余对象，组建最简决策表的算法。把求最简决策表的算法与高学东的粗集约简算法结合起来，形成了一种具有较好的处理不相容决策表、决策规则较合理的粗集约简算法。实例证明，该算法性能时间与空间复杂度与高学东算法相同，但明显优于Hu Xi-aoHua、叶东毅等的算法。
決策錶中大量的不相容對象和相同對象會降低粗集約簡算法性能，齣現錯誤的屬性覈和約簡結果。因此文中首先討論決策錶的不相容對象與相同對象即冗餘對象對約簡結果的影響，然後提齣利用對象頻度概唸來刪除冗餘對象，組建最簡決策錶的算法。把求最簡決策錶的算法與高學東的粗集約簡算法結閤起來，形成瞭一種具有較好的處理不相容決策錶、決策規則較閤理的粗集約簡算法。實例證明，該算法性能時間與空間複雜度與高學東算法相同，但明顯優于Hu Xi-aoHua、葉東毅等的算法。
결책표중대량적불상용대상화상동대상회강저조집약간산법성능，출현착오적속성핵화약간결과。인차문중수선토론결책표적불상용대상여상동대상즉용여대상대약간결과적영향，연후제출이용대상빈도개념래산제용여대상，조건최간결책표적산법。파구최간결책표적산법여고학동적조집약간산법결합기래，형성료일충구유교호적처리불상용결책표、결책규칙교합리적조집약간산법。실예증명，해산법성능시간여공간복잡도여고학동산법상동，단명현우우Hu Xi-aoHua、협동의등적산법。
Incompatible objects and the same objects in the decision table can degrade the performance of rough set reduction algorithms, producing wrong attribute core and reduction results. Therefore firstly it is discussed that incompatible objects and the same objects also called redundancy objects of the decision table affect reduction results. Secondly,objects frequency is put forward to remove redundant objects,the simplest decision table algorithm has been proposed. The algorithm for the simplest decision table is integrated with Gao Xue-dong's rough reduction algorithm,which could better deal with inconsistent decision tables and decision rules is more reasonable. Exam-ple proves that the performance time and space complexity of this algorithm is same as Gao Xuedong's algorithm,and is better than Hu Xiaohua's、Ye Dongyi's algorithms.
