摘要
拟周期性能描述对象在生命周期中重复性的趋势和走向 ,并能忽略时间轴上不规则的伸缩和幅度上的干扰 .该文以基于 Hbase分史制的 Web数据拟周期采掘任务为背景 ,提出了属性趋势、趋势惯量和峰谷链、抗干扰的惯性趋势算法和峰谷算法 ,对拟周期采掘给出一种解决方法 ,通过在一组地震数据上的采掘测试表明 。
The quasi periodicity describes the repeated behavior of specific objects while allowing uneven stretch or shrink on time axis, limited noises, and inflation /deflation of attribute values. To discover quasi periodicity from Web data, this work proposes the method to collect information from web sites. and the concepts of attribute tendency, tendency inertia and peak valley chain, as well as the anti noise algorithm with inertia, and peak valley algorithm, The article gives a solution to the problem of small noise and uneven stretched time axis in mining of quasi periodicity. The testing on a group of earthquake data shows that the method is useful and efficient.
出处
《计算机学报》
EI
CSCD
北大核心
2000年第1期52-59,共8页
Chinese Journal of Computers
基金
国家自然科学基金!( 69773 0 5 1)
关键词
Web知识发现
拟周期
时态数据库
数据采掘
Web KDD,quasi periodicity,temporal database,attribute tendency,tendency inertia
