Privileged Information for Data Clustering

TL;DR

This paper explores the use of privileged information in unsupervised data clustering, proposing new methods to improve stability and accuracy, and demonstrating their effectiveness on artificial and real-world datasets.

Contribution

It introduces the aRi-MAX and P-Dot algorithms that incorporate privileged information into clustering, extending Vapnik's supervised learning ideas to unsupervised settings.

Findings

aRi-MAX improves KMeans stability on artificial data

P-Dot fuses privileged and technical data for better clustering

Application to digit recognition confirms effectiveness

Abstract

Many machine learning algorithms assume that all input samples are independently and identically distributed from some common distribution on either the input space X, in the case of unsupervised learning, or the input and output space X x Y in the case of supervised and semi-supervised learning. In the last number of years the relaxation of this assumption has been explored and the importance of incorporation of additional information within machine learning algorithms became more apparent. Traditionally such fusion of information was the domain of semi-supervised learning. More recently the inclusion of knowledge from separate hypothetical spaces has been proposed by Vapnik as part of the supervised setting. In this work we are interested in exploring Vapnik's idea of master-class learning and the associated learning using privileged information, however within the unsupervised setting. Adoption of the advanced supervised learning paradigm for the unsupervised setting instigates investigation into the difference between privileged and technical data. By means of our proposed aRi-MAX method stability of the KMeans algorithm is improved and identification of the best clustering solution is achieved on an artificial dataset. Subsequently an information theoretic dot product based algorithm called P-Dot is proposed. This method has the ability to utilize a wide variety of clustering techniques, individually or in combination, while fusing privileged and technical data for improved clustering. Application of the P-Dot method to the task of digit recognition confirms our findings in a real-world scenario.