Representative Selection in Non Metric Datasets
This addresses the challenge of representative selection in domains like music and motion analysis where data lacks metric properties, offering a novel method for an existing bottleneck.
The paper tackles the problem of selecting a representative subset from non-metric datasets, where only pairwise similarities are available, by proposing δ-medoids as an extension to k-medoids, and empirically validates it in music and motion analysis with theoretical performance bounds.
This paper considers the problem of representative selection: choosing a subset of data points from a dataset that best represents its overall set of elements. This subset needs to inherently reflect the type of information contained in the entire set, while minimizing redundancy. For such purposes, clustering may seem like a natural approach. However, existing clustering methods are not ideally suited for representative selection, especially when dealing with non-metric data, where only a pairwise similarity measure exists. In this paper we propose $δ$-medoids, a novel approach that can be viewed as an extension to the $k$-medoids algorithm and is specifically suited for sample representative selection from non-metric data. We empirically validate $δ$-medoids in two domains, namely music analysis and motion analysis. We also show some theoretical bounds on the performance of $δ$-medoids and the hardness of representative selection in general.