数据挖掘与组合学习  

Data Mining and Ensemble of Learning Methods

作　　者：刁力力[1] 胡可云[1] 陆玉昌[1] 石纯一[1] 

机构地区：[1]清华大学计算机科学与技术系智能技术与系统国家重点实验室,北京100084

出　　处：《计算机科学》2001年第7期73-78,共6页Computer Science

基　　金：中国国家自然科学基金重大项目(编号:79990580);中国国家重点基础研究发展项目(编号:G1998030414);清华大学信息学院基础研究项目资助

摘　　要：Data-mining is a kind of solution for solving the problem of information exploding. Classification and prediction belong to the most fundamental tasks in data-mining field. Many experiments have showed that the results of ensemble of learning methods are generally better than those of single learning methods under most of the time. In the sense,it is of great value to introduce ensemble of learning methods to data mining. This paper introduces data mining and ensemble of learning methods respectively,along with the analysis and formulation about the role ensemble of learning methods can act in some important practicing aspects of data mining:Text mining,multi-media information mining and web mining.Data-mining is a kind of solution for solving the problem of information exploding. Classification and prediction belong to the most fundamental tasks in data-mining field. Many experiments have showed that the results of ensemble of learning methods are generally better than those of single learning methods under most of the time. In the sense.it is of great value to introduce ensemble of learning methods to data mining. This paper introduces data mining and ensemble of learning methods respectively, along with the analysis and formulation about the role ensemble of learning methods can act in some important practicing aspects of data mining:Text mining.multi-media information mining and web mining.

关 键 词：数据挖掘 数据库 知识发现 组合学习 

分 类 号：TP311.13[自动化与计算机技术—计算机软件与理论]