数据挖掘与组合学习

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.

关键词：Data-mining、Classification/prediction、Ensemble of learning methods

所属期刊栏目：28

分类号：TP3(计算技术、计算机技术)

资助基金：国家自然科学基金79990580;国家重点基础研究发展计划973计划G1998030414;清华大学校科研和教改项目

在线出版日期：2004-01-08（万方平台首次上网日期，不代表论文的发表时间）

页数：共6页

页码：73-78