Semi-supervised spectral clustering with automatic propagation of pairwise constraints

Abstract

International audienceIn our data driven world, clustering is of major importance to help end-users and decision makers understanding information structures. Supervised learning techniques rely on ground truth to perform the classification and are usually subject to overtraining issues. On the other hand, unsupervised clustering techniques study the structure of the data without disposing of any training data. Given the difficulty of the task, unsupervised learning tends to provide inferior results to supervised learning. To boost their performance, a compromise is to use learning only for some of the ambiguous classes. In this context, this paper studies the impact of pairwise constraints to unsupervised Spectral Clustering. We introduce a new generalization of constraint propagation which maximizes partitioning quality while reducing annotation costs. Experiments show the efficiency of the proposed scheme

Deep learning vs spectral clustering into an active clustering with pairwise constraints propagation

“…In this case, only one pair is randomly selected from all the not annotated pairs and is submitted to the Oracle. Next, the automatic constraint propagation described in [4] is applied. This step guaranties that the maximum number of supervision loops to perform equals the number of considered individuals of the dataset.…”

“…In view of this idea, it is of major interest to optimize the constraints (i.e., annotated pairs) thus to maximize clustering quality while minimizing the costs of user knowledge acquisition. One of the most common strategies consists in using a pairwise constraint automatic propagation approach [4]. This will be one of the two aspects of this paper.…”