Semi-supervised Clustering: Learning with Limited User Feedback
Sugato Basu · 2004
In many machine learning domains (e.g. text processing, bioinformatics), there is a large supply of unlabeled data but limited labeled data, which can be expensive to generate. Consequently, semi-supervised learning, learning from a combination of both labeled and unlabeled data, has become a topic of significant recent interest. In the proposed thesis, our research focus is on semi-supervised clustering, which uses a small amount of supervised data in the form of class labels or pairwise constraints on some examples to aid unsupervised clustering. Semi-supervised clustering can be either search-based, i.e., changes are made to the clustering objective to satisfy user-specified labels/constraints, or similarity-based, i.e., the clustering similarity metric is trained to satisfy the given labels/constraints. Our main goal in the proposed thesis is to study search-based semi-supervised clustering algorithms and apply them to different domains. In our initial work, we have shown how supervision can be provided to clustering in the form of labeled data points or pairwise constraints. We have also developed an active learning framework for selecting informative constraints in the pairwise constrained semi-supervised clustering model, and proposed a method for unifying searchbased and similarity-based techniques in semi-supervised clustering. In this thesis, we want to study other aspects of semi-supervised clustering. Some of the issues we want to investigate include: (1) effect of noisy, probabilistic or incomplete supervision in clustering; (2) model selection