Generalized Dirichlet-process-means for f-separable distortion measures
Generalized Dirichlet-process-means for f-separable distortion measures
复制标题
f-可分离失真测量的广义狄利克雷过程均值
DOI：
10.1016/j.neucom.2020.03.123
复制
发表时间：
2021
期刊：
影响因子：
6
通讯作者：
Kazuho Watanabe
中科院分区：
文献类型：
--
作者：
Masahiro Kobayashi;Kazuho Watanabe
DP-均值聚类是K-均值聚类的一种扩展。虽然它是一个简单而有效的算法实现，它可以同时估计集群的数量。然而，DP均值是专门为平均失真度量而设计的。因此，它很容易受到离群数据，并可能导致大的最大失真的集群。在这项工作中，我们扩展的DP-手段的f-可分离的失真措施的目标函数，并提出了一个统一的学习算法，以克服上述问题，通过选择函数f。此外，估计的聚类中心的影响函数进行了分析，以评估对离群值的鲁棒性。我们使用真实的数据集的数值实验证明了广义方法的性能。
DP-means clustering was obtained as an extension ofK-means clustering. While it is implemented with a simple and efficient algorithm, it can estimate the number of clusters simultaneously. However, DP-means is specifically designed for the average distortion measure. Therefore, it is vulnerable to outliers in data, and can cause large maximum distortion in clusters. In this work, we extend the objective function of the DP-means tof-separable distortion measures and propose a unified learning algorithm to overcome the above problems by selecting the functionf. Further, the influence function of the estimated cluster center is analyzed to evaluate the robustness against outliers. We demonstrate the performance of the generalized method by numerical experiments using real datasets.