摘要
该文依据供水收费管理系统中的收费欠费问题,利用决策树和频繁项集相结合的方法对供水收费数据进行处理,挖掘影响营业厅供水收费时产生违约金的因素。通过计算每个属性的信息增益以及优化的频繁项集挖掘出了影响收费的主要因子。结果表明两种方法得出的结论一致,两种方法的结合使用使得结果更精确更有效。为供水收费,减少违约金,提供了可靠的决策支持。
This paper, based on the Water Charge Management System, uses the method of combines the decision tree with the frequent itemsets, excavates the factors that influences the produce of overdue fine. By means of calculating every property's information gain and optimized frequent itemsets, it excavates the main factors that influence charge. The results show that the two methods get the same conclusion, the combination of this two methods make the result more accurate and more effective. It provides reliable decision support to the charge of Water Charge Management System.
出处
《电脑知识与技术(过刊)》
2016年第6X期1-4,共4页
Computer Knowledge and Technology
关键词
决策树
频繁项集
违约金
影响因子
动态分析
decision tree
frequent itemsets
overdue fine
impact factor
dynamic analysis