【摘要】 分类和聚类是数据挖掘中两个重要的研究领域,分类需要相关的先验知识,而聚类往往依据某种相似性测度,从数据本身来寻找其内在特征.在电力系统负荷预测过程中,依靠先验知识得到的分类结果与聚类结果之间并不协调.针对这一问题,文中给出了调和矩阵的定义,并在此基础上,提出调和聚类-分类算法,将该方法应用于电力系统负荷预测的样本分类中,实际结果表明,通过文中方法得到的分类结果更加客观和科学,预测结果的可靠性得到了保证.更多还原

【Abstract】 Clustering and classification are two important research areas of data mining.Classification needs related prior-knowledge,while clustering normally finds its own inherent characteristics from the data based on similarity measure.In the process of power load forecasting,the results of classification and clustering are inconsistent.For this problem,this paper propose the definition of associated matrix and on that basis propose associated clustering-classification algorithm.This algorithm is applied to data sample classification for power load prediction,the experiment show that the classification results obtained by our method are more reliable.更多还原