• 主题
高级搜索  >
历史搜索 清空历史
首页> 外文会议>International Conference on Cloud Computing and Security >An Improved Artificial Immune Network Based on the Secondary Immune Mechanism for Data Clustering
【24h】

An Improved Artificial Immune Network Based on the Secondary Immune Mechanism for Data Clustering

机译：一种基于二级免疫机制的改进的人工免疫网络数据聚类

获取原文
页面导航
  • 摘要
  • 著录项
  • 相似文献
  • 相关主题

摘要

Data clustering is a typical method in data mining. As a effective algorithm for clustering, the Artificial Immune Network is inspired by natural immune system can reflect the structure of the given dataset, filter redundancy and cluster datasets without the number of clusters, so far it is widely used. However, it can't effectively identify the noise nodes, the running time is long and too much parameters are set in improved algorithms. In order to shorten running time and reduce the impact of parameters, this paper proposes an improved artificial immune network based on the secondary immune mechanism. The Clone operator and Mutation operator are replaced by Competition Selection operator and Competition Selection strategy, which are inspired by the resource limited artificial immune system. Because the algorithm can reach a stable convergence only through two times, so it greatly reduce the running time; and can effectively identify the noise nodes due to the introduction of stimulation level. A number of datasets including artificial datasets and real-world datasets are used to evaluate the performance of the proposed algorithm and the other existing clustering algorithms, such as K-means, FCM, SC, aiNet and FCAIN. The simulation results indicate that the proposed artificial immune network algorithm is an effective and efficient method in data clustering.
机译：数据聚类是数据挖掘中的一种典型方法。作为一种有效的聚类算法，人工免疫网络受到自然免疫系统的启发，可以反映给定数据集的结构，过滤器冗余和聚类数据集而无需聚类数量，到目前为止已被广泛使用。但是，它不能有效地识别噪声节点，运行时间长，改进算法中设置的参数过多。为了缩短运行时间并减少参数的影响，本文提出了一种基于二级免疫机制的改进的人工免疫网络。克隆运算符和变异运算符被竞争选择运算符和竞争选择策略所取代，这受到资源有限的人工免疫系统的启发。由于该算法只能经过两次才能达到稳定收敛，因此大大减少了运行时间;由于引入了刺激水平，可以有效地识别噪声节点。许多数据集（包括人工数据集和现实世界数据集）用于评估所提出算法和其他现有聚类算法（例如K-means，FCM，SC，aiNet和FCAIN）的性能。仿真结果表明，提出的人工免疫网络算法是一种有效的数据聚类方法。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱：kefu@zhangqiaokeyan.com

京公网安备：11010802029741号 ICP备案号：京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有

  • 客服微信

  • 服务号