【摘要】 时间的描述和划分是时态数据采掘中一个非常重要的方面,针对目前时态数据采掘中缺少对多粒度时间等的研究的现状,提出了多粒度时间,粒度转换,时态序等的严格数学定义,并研究和证明了它们的相关性质。以此为基础引出了一个多粒度时间部分周期模型,对模型的支持度和置信度等性质进行了详细讨论,并将多粒度时间下的部分周期模型运用到股票数据实验中,实验表明所提出的模型对于研究时态数据采掘具有重要意义。更多还原

【Abstract】 The description and division of time is a very important aspect;For resolving the problem of lacking research of multiple time granularity at present,a mathematical concepts of multiple time granularity,granularity transform and temporal series is presented.The relationships and properties of model is proved.A multiple time granularity partial periodicity model is introduced based on those.At the same time,the support and confidence properties of model is discussed and is applied to the stock data experiment,which is significant in study of the temporal data mining.更多还原