基于C4．5和Boosting算法的数据库负载自动识别

Automatic recognition of database load based on C4.5 and boosting algorithm

对数据库进行调整与优化时,负载的类型是要考虑的一个关键因素。不同的负载类型(联机事务处理OLTP和联机分析处理OLAP),意味着不同的资源分配策略。另外在系统运行中,负载的类型是经常变化的。把基于C4.5决策树算法和Bo- osting集成学习算法的分类模型用于数据库负载类型的自动识别中,实验结果表明,该分类模型是可行的,达到了令人满意的准确度以及关于负载识别的4个特殊要求。

The type of the load is a key consideration in tuning the database system. The strategies for resource allocation are different depending on whether it is online transactional processing （OLTP） or online analytical processing （OLAP）, In addition, the type of load often changes during the normal processing cycle, A classification model based on C4.5 and boosting algorithm is proposed to auto- matically recognize the type of the load. The experimental results show that the classifiers are reliable, have satisfactory accuracy and meet the four special requirements for automatic recognition of database load.