Study on The Outlier Detection Algorithm Based on Clustering Reduction
The computation complexity of the algorithm for identifying density-based local outliers (LOF algorithm) is not ideal,which affects its applications in large scale data sets. Under such circumstances,an outlier detection algorithm based on kernel k-means clustering was proposed,which applied kernel k-means clustering on data sets to calculate representation degrees (rpd ) of clusters. Those clustering data sets with high rdp were carried out for the candidates of outliers which were left for the following outlier mining using LOF algorithm. The outlier detection algorithm based on kernel k-means clustering reduced the computation for neighborhood of the data sets,and shortens the execution time. Result in both theoretic analyses and experiments show that this algorithm is effective and efficient.
data mining outlier detection Kernel K-means clustering representation degrees
Zhang Yuanfang Qin Liangxi
School of Computer,Electronics and Information,Guangxi University,Nanning 530004,China
国际会议
南宁
英文
152-156
2010-12-10（万方平台首次上网日期，不代表论文的发表时间）