摘要
探索性仿真是一种研究复杂系统的科学手段,而数据挖掘是处理探索性仿真所产生的海量数据,而在实施数据挖掘前如何进行有效的数据预处理成为当前仿真领域面临的难题。为解决目前在探索性仿真中数据预处理工作存在的目标不够明确、重点不够突出等问题,提出了探索性仿真数据预处理需求分析,结合探索性仿真数据的特点,首先分析了通用数据预处理需求,之后对决策树挖掘、关联规则挖掘、聚类分析三种典型数据挖掘算法的数据预处理需求进行了分析。研究成果较好地满足了探索性仿真数据预处理工作的需求。
Exploratory simulation experiment is a scientific method to research complicated system, while data mining is an important method to process a great deal of data produced in an exploratory simulation experiment. How to preprocess data effectively before data mining is a problem to be solved for the domestic simulation realm. In order to solve the problem of the ambiguity of goal and unapparent emphasis of data preprocessing, requirement analysis of data preprocessing in exploratory simulation experiment was put forward. Combing with the specialty of data produced in exploratory simulation experiment, the general requirement of data preprocessing was analyzed in the paper firstly. The requirements of the algorithms of decision tree, association rules, and cluster analysis of data preprocessing were analyzed then. The research results of this paper can meet the requirement of data preprocessing in exploratory simu- lation experiment.
出处
《计算机仿真》
CSCD
北大核心
2012年第11期64-67,共4页
Computer Simulation
关键词
探索性仿真
数据预处理
需求分析
Exploratory simulation
Data preprocessing
Requirement analysis
