A Sequence Classification Model Based on Pattern Coverage Rate
暂无分享,去创建一个
The technique of classification can sort data into various categories for data mining studies. The demand for sequence data classification has increased with the development of information technology. Several applications involve decision prediction based on sequence data, but the traditional classification methods are unsuitable for sequence data. Thus, this paper proposes a Pattern Coverage Rate-based Sequence Classification Model (PCRSCM) to integrate sequential pattern mining and classification techniques. PCRSCM mines sequential patterns to find characteristics of each class, and then calculates pattern coverage rates and class scores to predict the class of a sequence. The experimental results show that PCRSCM exhibits excellent prediction performance on synthetic and real sequence data.
[1] Dimitrios I. Fotiadis,et al. A two-stage methodology for sequence classification based on sequential pattern mining and optimization , 2008, Data Knowl. Eng..
[2] Anne Laurent,et al. S2MP: Similarity Measure for Sequential Patterns , 2008, AusDM.
[3] Jian Pei,et al. Debt Detection in Social Security by Sequence Classification Using Both Positive and Negative Patterns , 2009, ECML/PKDD.
[4] Dimitris G. Papageorgiou,et al. MERLIN-3.1.1. A new version of the Merlin optimization environment , 2004 .