Semi-supervised clustering with limited background knowledge
Sugato Basu · 2004
In many machine learning domains, there is a large supply of unlabeled data but limited labeled data, which can be expen-sive to generate. Consequently, semi-supervised learning, learning from a combination of both labeled and unlabeled data, has become a topic of significant recent interest. 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 unsuper-vised clustering. Semi-supervised clustering can be either constraint-based, i.e., changes are made to the clustering ob-jective to satisfy user-specified labels/constraints, or metric-based, i.e., the clustering distortion measure is trained to sat-isfy the given labels/constraints. Our main goal in this thesis is to study constraint-based semi-supervised clustering algo-rithms, integrate them with metric-based approaches, char-