Fuzzy Clustering Algorithm based on Factor Analysis and its Application to Mail Filtering

摘要

Aim at the faults of Dynamic ClusteringAlgorithm based on Fuzzy Equation Matrix, we raise afuzzy clustering algorithm based on factor analysis, which itcombines the technology of reducing dimension using factoranalyses method. The algorithm will deal with the samplecollections before fuzzy clustering, which enlarge the scaleof using dynamic clustering algorithm to resolve practicalproblems. All these show that the algorithm has a strongcapability of concluding and abstracting through beingapplied to E-mail filtering. At the same time, we also makean experiment in our optional database. The experimentresult verifies that the algorithm recall rate is 87.3 % in themail filtering, which is higher than the SVM’s 80.1%, Na?veBayes’s 61.7%, and KNN’s 73.2% respectively. Theexperiments show that the new algorithm has better recallrate and error rate.