Towards semi-supervised ensemble clustering using a new membership similarity measure

Hierarchical clustering is a common type of in which the dataset hierarchically divided and represented by dendrogram. Agglomerative Clustering (AHC) hierarchical clusters are created bottom-up. In addition, semi-supervised new method field machine learning, where supervised unsupervised learning combined. performance effectively improved as it uses small amount labelled data to aid learning. Meanwhile, ensemble combining results several individual methods can achieve better compared each methods. Considering AHC with for configuration has received less attention past literature. order results, we propose framework developed based on AHC-based Here, develop flexible weighting mechanism along membership similarity measure that establish compatibility between We evaluated proposed equivalent wide variety UCI datasets. Experimental show effectiveness from different aspects such NMI, ARI accuracy.