|
|||||
|
|
| 基于动态阻尼因子的动态阻尼仿射传播聚类 | |
| 引用本文： | 王羡慧,陈俊英,覃征,张选平.基于动态阻尼因子的动态阻尼仿射传播聚类[J].华中科技大学学报(自然科学版),2012,40(2):116-122. |
| 作者姓名： | 王羡慧 陈俊英 覃征 张选平 |
| 作者单位： | 1. 西安交通大学电子与信息工程学院,陕西西安710049／新疆大学信息科学与工程学院,新疆乌鲁木齐830046 2. 西安建筑科技大学信息与控制工程学院,陕西西安,710055 3. 西安交通大学电子与信息工程学院,陕西西安710049／清华大学计算机科学与技术系,北京100084 4. 西安交通大学电子与信息工程学院,陕西西安,710049 |
| 基金项目： | 国家自然科学基金资助项目，高等学校博士学科点专项科研基金资助项目，西安建筑科技大学人才科技基金资助项目 |
| 摘 要： | 为了有效地控制仿射传播聚类的全局搜索和局部搜索，将仿射传播聚类视作一个搜索能量函数最小值的方法，提出一种基于动态阻尼因子策略的动态阻尼仿射传播聚类．使用两阶段动态阻尼因子策略，使得仿射传播聚类迭代过程中阻尼因子不再是静态固定值，而是自适应的动态改变值，构造了1种线性策略和2种非线性策略来动态改变阻尼因子的值．在标准数据集上的实验结果表明：动态阻尼仿射传播聚类避免了仿射传播聚类发生震荡，同时可以加快仿射传播聚类的收敛． |
| 关 键 词： | 仿射传播聚类 动态阻尼因子 线性策略 凸函数策略 凹函数策略 |
Dynamic damping factor-based dynamic damping affinity propagation clustering |
|
| Wang Xianhui,Chen Junying,Qin Zheng,Zhang Xuanping.Dynamic damping factor-based dynamic damping affinity propagation clustering[J].JOURNAL OF HUAZHONG UNIVERSITY OF SCIENCE AND TECHNOLOGY.NATURE SCIENCE,2012,40(2):116-122. | |
| Authors: | Wang Xianhui Chen Junying Qin Zheng Zhang Xuanping |
| Institution: | 1(1 School of Electronics and Information Engineering,Xi′an Jiaotong University,Xi′an 710049, China;2 School of Information Science and Engineering,Xinjiang University,Urumqi 830046,China; 3 School of Information and Control Engineering,Xi′an University of Architecture and Technology,Xi′an 710055,China;4 Department of Computer Science and Technology,Tsinghua University,Beijing 100084,China) |
| Abstract: | Affinity propagation clustering(APC) can be viewed as a method that searches for minima of an energy function.To control the global and local search of APC efficiently,dynamic damping factor affinity propagation clustering based on dynamic damping factor strategy was proposed,which used a two-stage dynamic damping factor strategy.The strategy made it true that the damping factor was no longer a static fixed value,but the adaptive value of the dynamic changes in the iterative process of APC.A linear strategy and two nonlinear strategies were constructed to dynamically change the damping factor’s value.The experimental results in standard data sets show that the dynamic damping factor affinity propagation clustering can not only avoid oscillations,but also significantly improve convergence rate. |
| Keywords: | affinity propagation clustering (APC) dynamic damping factor linear strategy convexfunction strategy concave function strategy |
| 本文献已被 CNKI 万方数据 等数据库收录！ | |