Robust semi-supervised clustering with polyhedral and circular uncertainty

A new Semi-Supervised Intuitionistic Fuzzy C-means Clustering

“…Blum et al [7] used graph based method to provide information regarding labeling in the process of unsupervised clustering. Dinler et al [8] have given the semi-supervised clustering algorithm that aims to partition regional data objects in the presence of instance level constraints. Saha et al [9] proposed the concept of semisupervised clustering using multiobjective optimization and applied the concept in the process of automatic medical image segmentation.…”

“…Lot of efforts are required by domain experts for the process of labelling the data whereas unlabelled data is cheap and available in abundance. Semi-supervised learning use limited label data to learn the input variable and make some initial predictions required for the unsupervised learning process [5]- [8].…”

“…A salient feature of partial supervision in the clustering algorithms is the availability of labelled data [5] or the presence of some constraints [8]. In this SSFCM model partial supervision is provided in the form of labels.…”

“…In semi-supervised clustering, a small set of labelled data is provided with the information to encounter some initial variable constraints, along with large set of unlabelled data. The amount of label data available is not sufficient for the process of clustering [8]. So X in case of SSFCM clustering can be represented as The membership matrix for unlabelled data u im  will be initialized randomly [13].…”