A comprehensive comparative study of clustering-based unsupervised defect prediction models
这篇论文将40个无监督模型分成了9个聚类族,帮助我(初学者)建立基于聚类的无监督模型的初步印象,然而缺乏直观的认识。在搜索无监督模型的过程中,我发现有一些十分生动的GIF图描述这些模型的聚类过程,因此,我打算将二者组合起来。
1.基于分区的聚类族/PBC/Partition Based Clustering family)
K-Means (Hartigan and Wong,1979), |
|
Cascade K-Means(CM) (Karegowda等人,2012), |
|
Canopy (McCallum等人,2000), |
|
X-Means(Peleg等人,2000), |
|
K-Medoids (Jin and Han,2016), |
|
围绕Medoids的划分(Partitioning Around Medoids / PAM)(Kaufman and Rousseuw,2009), |
|
Mini Batch K-Means(MBM),(Alonso,2013), |
|
Fuzzy C-Means (FCM) (Bezdek et al., 1984), |
|
Fuzzy C-Shell (FCS) (Dave, 1990), |
|
Hard C-Means (HCM) (MacQueen et al., 1967), |
|
K-Modes (Huang, 1997), |
|
FarthesFirst (FF) (Hochbaum and Shmoys, 1985), |
|
Clustering LARge Applications (CLARA) (Kaufman and Rousseeuw, 2009). |
|