摘要
基于粗糙集理论的启发式数据挖掘算法,先量化数据属性的重要性,再对属性排序,并对选择的属性子集进行约简条件判定。在选择若干次后获得满足约简条件的属性子集。利用该算法,通过预处理、数据约减等步骤从商场营销数据库中收集顾客属性数据,消去多余数据属性。并由此预测需求趋势,评估需求倾向的改变,以便进行决策。
Heuristic data mining algorithm based on rough set theory quantified attribute importance of data, then classified the attributes and judged reduction condition about the selected attribute subset. After a number of repeats above, it gains the attribute subset, which satisfied with reduction condition. Using the algorithm, system can collect customers＇ attributes data and eliminate redundant data attributes from emporium marketing database by pre（reatment, data reduction and other steps. According to this predicts requirement trend, evaluates the changing of requirement tendency for decision-making.
出处
《兵工自动化》
2006年第6期37-38,共2页
Ordnance Industry Automation
基金
甘肃省自然基金项目(3ZS042-B25-038)
甘肃省建设科技计划项目(JK2005-21)
关键词
粗糙集
数据挖掘
属性约简
属性重要性
营销数据库
Rough set
Data mining
Attribute reduction
Attribute importance
Marketing database
