正在加载图片...
第11期 王玲等:一种基于密度的模糊自适应聚类算法 ·1565· [4]Shi H Y,Ping L,Ji D G,et al.A statistical information-based 4结论 clustering approach in distance space.I Zhejiang Unir Sci A, 2005,6(1):71 DFAC算法通过自适应地确定邻域半径得到每 [5]Zhao W,Xia G S,Gou Z J,et al.An Improved DBSCAN Algo- 个样本的密度,并基于样本密度逐渐增加聚类中心, rithm.J Sichuan Norm Unin Nat Sci,2013(2):312 同时为了保证聚类结果的有效性提出一种新的模糊 (赵文,夏桂书,苟智坚,等.一种改进的DBSCAN算法.四 聚类有效性指标以判断聚类数为何值时聚类效果最 川师范大学学报:自然科学版,2013(2):312) 佳.最终对聚类中心的更新进一步改善了算法的聚 6 Xia L N,Jing J W.SA-DBSCAN:a self-adaptive densitybased 类正确率.真实数据实验结果表明,DFAC算法无需 clustering algorithm.J Grad Sch Chin Acad Sci,2009,26(4): 530 预先定义聚类邻域参数,在未知聚类数的情况下,根 (夏鲁宁,荆继武.SA-DBSCAN:一种自适应基于密度聚类 据新的模糊聚类指标可以确定最佳聚类数,因此极 算法.中国科学院研究生院学报,2009,26(4):530) 大提高了算法的自适应性,而且DFAC算法的准确 7]Huang M,Bian F.A grid and density based fast spatial clustering 性和效率较DBSCAN算法有显著的提高. algorithm /AICl'09 International Conference on Artificial Intelli- gence and Computational Intelligence.IEEE,2009:260 参考文献 8] Wang W,Li D Z,Vrbanek J.An evolving neuro-fuzzy technique [Ertoz L,Steinbach M,Kumar V.Finding clusters of different si- for system state forecasting.Neurocomputing,2012,87:111 zes,shapes,and densities in noisy high dimensional data//Siam Bai L.Theoretical Analysis and Effective Algorithms of Cluste Proceedings Series,2003:47 Learning [Dissertation].Taiyuan:Shanxi University,2012 Ester M,Kriegel H P,Sander J,et al.A density-based algorithm (白亮.聚类学习的理论分析与高效算法研究[学位论文」 for discovering clusters in large spatial databases with noise / 太原:山西大学,2012) Proceedings of the 2nd International Conference on Knowledge Dis- [10]Bezdek JC.Ehrlich R.Full W.FCM:the fuzzy c-means cluste- covery and Data Mining.Portland,1996:226 ring algorithm.Comput Geosci,1984,10(2):191 Lin C Y,Chang CC,Lin CC.A new density-based scheme for [11]Sun Y,Zhu Q M,Chen Z X.An iterative initial-points refine- clustering based on genetic algorithm.Fundam Inf,2005,68 ment algorithm for categorical data clustering.Pattern Recognit (4):315 Lt,2002,23(7):875第 11 期 王 玲等: 一种基于密度的模糊自适应聚类算法 4 结论 DFAC 算法通过自适应地确定邻域半径得到每 个样本的密度,并基于样本密度逐渐增加聚类中心, 同时为了保证聚类结果的有效性提出一种新的模糊 聚类有效性指标以判断聚类数为何值时聚类效果最 佳. 最终对聚类中心的更新进一步改善了算法的聚 类正确率. 真实数据实验结果表明,DFAC 算法无需 预先定义聚类邻域参数,在未知聚类数的情况下,根 据新的模糊聚类指标可以确定最佳聚类数,因此极 大提高了算法的自适应性,而且 DFAC 算法的准确 性和效率较 DBSCAN 算法有显著的提高. 参 考 文 献 [1] Ertz L,Steinbach M,Kumar V. Finding clusters of different si￾zes,shapes,and densities in noisy high dimensional data / / Siam Proceedings Series,2003: 47 [2] Ester M,Kriegel H P,Sander J,et al. A density-based algorithm for discovering clusters in large spatial databases with noise / / Proceedings of the 2nd International Conference on Knowledge Dis￾covery and Data Mining. Portland,1996: 226 [3] Lin C Y,Chang C C,Lin C C. A new density-based scheme for clustering based on genetic algorithm. Fundam Inf,2005,68 ( 4) : 315 [4] Shi H Y,Ping L,Ji D G,et al. A statistical information-based clustering approach in distance space. J Zhejiang Univ Sci A, 2005,6( 1) : 71 [5] Zhao W,Xia G S,Gou Z J,et al. An Improved DBSCAN Algo￾rithm. J Sichuan Norm Univ Nat Sci,2013( 2) : 312 ( 赵文,夏桂书,苟智坚,等. 一种改进的 DBSCAN 算法. 四 川师范大学学报: 自然科学版,2013( 2) : 312) [6] Xia L N,Jing J W. SA-DBSCAN: a self-adaptive density-based clustering algorithm. J Grad Sch Chin Acad Sci,2009,26 ( 4) : 530 ( 夏鲁宁,荆继武. SA--DBSCAN: 一种自适应基于密度聚类 算法. 中国科学院研究生院学报,2009,26( 4) : 530) [7] Huang M,Bian F. A grid and density based fast spatial clustering algorithm / / AICI’09 International Conference on Artificial Intelli￾gence and Computational Intelligence. IEEE,2009: 260 [8] Wang W,Li D Z,Vrbanek J. An evolving neuro-fuzzy technique for system state forecasting. Neurocomputing,2012,87: 111 [9] Bai L. Theoretical Analysis and Effective Algorithms of Cluster Learning [Dissertation〗. Taiyuan: Shanxi University,2012 ( 白亮. 聚类学习的理论分析与高效算法研究[学位论文〗. 太原: 山西大学,2012) [10] Bezdek J C,Ehrlich R,Full W. FCM: the fuzzy c-means cluste￾ring algorithm. Comput Geosci,1984,10( 2) : 191 [11] Sun Y,Zhu Q M,Chen Z X. An iterative initial-points refine￾ment algorithm for categorical data clustering. Pattern Recognit Lett,2002,23( 7) : 875 · 5651 ·
<<向上翻页
©2008-现在 cucdc.com 高等教育资讯网 版权所有