CAJ | 학술논문

在大型流媒体数据库数据集中，交叉性数据的鲁棒性挖掘结构建立是实现对数据库差异属性分类和数据访问的基础。传统方法对大型数据库中的交叉性数据的鲁棒性挖掘结构建模采用基于遗传算法的数据集聚调度方法，存在较大的路径损耗，数据挖掘的鲁棒性不好。提出改进的基于局部离群点检测遗传进化的大型数据库交叉型数据挖掘模型，构建基于遗传算法的大型流媒体数据库挖掘结构，进行大型流媒体数据库中交叉型数据信息流特征预处理，结合交叉性型数据的离群因子概念，调整流媒体数据调度的位置变换策略，实现交叉性数据的鲁棒性挖掘算法改进。仿真实验结果表明，该算法能有效数据挖掘的a最大匹配率和局部离群点检测性能，保证了数据挖掘的鲁棒性，各项参数指标优于传统方法，展示了较好的应用价值。
재대형류매체수거고수거집중，교차성수거적로봉성알굴결구건립시실현대수거고차이속성분류화수거방문적기출。전통방법대대형수거고중적교차성수거적로봉성알굴결구건모채용기우유전산법적수거집취조도방법，존재교대적로경손모，수거알굴적로봉성불호。제출개진적기우국부리군점검측유전진화적대형수거고교차형수거알굴모형，구건기우유전산법적대형류매체수거고알굴결구，진행대형류매체수거고중교차형수거신식류특정예처리，결합교차성형수거적리군인자개념，조정류매체수거조도적위치변환책략，실현교차성수거적로봉성알굴산법개진。방진실험결과표명，해산법능유효수거알굴적a최대필배솔화국부리군점검측성능，보증료수거알굴적로봉성，각항삼수지표우우전통방법，전시료교호적응용개치。
In the large scale streaming media data of database, the robustness of the cross data mining structure built is the foundation to realize database attribute classification and data access. The traditional method used genetic algorithm sched?uling method for large database mining based on data gathering, the path loss greater robustness is occurred, data mining is not good. An improved large-scale streaming media database structure mining algorithm is proposed based on local outlier large database cross type data point detection and genetic evolution, crossover type data information flow feature prepro?cessing of the large-scale media streaming database is constructed, combined with cross type data outlier factor concept, po?sition transformation strategy is adjusted, streaming media data scheduling is obtained to achieve the robustness, improved cross data mining algorithm. Simulation results show that the algorithm can improve the maximum matching rate, local outli?er detection performance is better, which guarantee the robustness of data mining. The parameters is better than the tradi?tional methods, it has a good application value.