Timezone: »
The goal of clustering is to group similar objects into meaningful partitions. This process is well understood when an explicit similarity measure between the objects is given. However, far less is known when this information is not readily available and, instead, one only observes ordinal comparisons such as ``object i is more similar to j than to k.'' In this paper, we tackle this problem using a two-step procedure: we estimate a pairwise similarity matrix from the comparisons before using a clustering method based on semi-definite programming (SDP). We theoretically show that our approach can exactly recover a planted clustering using a near-optimal number of passive comparisons. We empirically validate our theoretical findings and demonstrate the good behaviour of our method on real data.
Author Information
Michaël Perrot (INRIA)
Pascal Esser (Technical University of Munich)
Debarghya Ghoshdastidar (Technical University Munich)
More from the Same Authors
-
2021 : Towards Modeling and Resolving Singular Parameter Spaces using Stratifolds »
Pascal Esser · Frank Nielsen -
2021 : Towards Modeling and Resolving Singular Parameter Spaces using Stratifolds »
Pascal Esser · Frank Nielsen -
2022 : Fairness Certificates for Differentially Private Classification »
Paul Mangold · Michaël Perrot · Marc Tommasi · Aurélien Bellet -
2021 : Contributed Talks in Session 2 (Zoom) »
Courtney Paquette · Chris Junchi Li · Jeffery Kline · Junhyung Lyle Kim · Pascal Esser -
2021 : Poster Session 1 (gather.town) »
Hamed Jalali · Robert Hönig · Maximus Mutschler · Manuel Madeira · Abdurakhmon Sadiev · Egor Shulgin · Alasdair Paren · Pascal Esser · Simon Roburin · Julius Kunze · Agnieszka Słowik · Frederik Benzing · Futong Liu · Hongyi Li · Ryotaro Mitsuboshi · Grigory Malinovsky · Jayadev Naram · Zhize Li · Igor Sokolov · Sharan Vaswani -
2021 Poster: Learning Theory Can (Sometimes) Explain Generalisation in Graph Neural Networks »
Pascal Esser · Leena Chennuru Vankadara · Debarghya Ghoshdastidar -
2019 Poster: Foundations of Comparison-Based Hierarchical Clustering »
Debarghya Ghoshdastidar · Michaël Perrot · Ulrike von Luxburg -
2018 Poster: Practical Methods for Graph Two-Sample Testing »
Debarghya Ghoshdastidar · Ulrike von Luxburg -
2016 Poster: Mapping Estimation for Discrete Optimal Transport »
Michaël Perrot · Nicolas Courty · Rémi Flamary · Amaury Habrard -
2015 Poster: Regressive Virtual Metric Learning »
Michaël Perrot · Amaury Habrard -
2014 Poster: Consistency of Spectral Partitioning of Uniform Hypergraphs under Planted Partition Model »
Debarghya Ghoshdastidar · Ambedkar Dukkipati