Spatial Clustering with Obstacles Constraints Using PSO-DV and K-Medoids
Spatial Clustering with Obstacles Constraints(SCOC)has been a new topic in Spatial Data Mining(SDM).Inthis paper,we propose an advanced Particle SwarmOptimization(PSO)and Differential Evolution(DE)methodfor SCOC.In the process of doing so,we first developed anovel spatial obstructed distance using PSO-DV(Particle SwarmOptimization with Differentially perturbed Velocity)based ongrid model to obtain obstructed distance,which is namedPDGSOD,and then we presented a new PDKSCOC based onPSO-DV and K-Medoids to cluster spatial data with obstaclesconstraints.The experimental results show that PDGSOD iseffective,and PDKSCOC can not only give attention to higherlocal constringency speed and stronger global optimum search,but also get down to the obstacles constraints and practicalitiesof spatial clustering;and it performs better than Improved K-Medoids SCOC(IKSCOC)in terms of quantization error andhas higher constringency speed than Genetic K-Medoids SCOC(GKSCOC).
Xueping Zhang Wei Ding Jiayao Wang Zhongshan Fan Gaofeng Deng
Information Science& Engineering Henan Univ.of Technology Zhengzhou,Henan China 450001;Key Laborator Information Science& Engineering Henan Univ.of Technology Zhengzhou,Henan China 450001 Information Science& Engineering Henan Univ.of Technology Zhengzhou,Henan China 450001;Surveying & M Henan Academy of Traffic Science & Technology Zhengzhou,Henan China 450052
国际会议
厦门
英文
246-251
2008-11-17（万方平台首次上网日期，不代表论文的发表时间）