Elsevier

Journal of Algorithms

Volume 36, Issue 2, August 2000, Pages 205-240
Journal of Algorithms

Regular Article
Parallel Algorithms for Hierarchical Clustering and Applications to Split Decomposition and Parity Graph Recognition

https://doi.org/10.1006/jagm.2000.1090Get rights and content

Abstract

We present efficient (parallel) algorithms for two hierarchical clustering heuristics. We point out that these heuristics can also be applied to solving some algorithmic problems in graphs, including split decomposition. We show that efficient parallel split decomposition induces an efficient parallel parity graph recognition algorithm. This is a consequence of the result of S. Cicerone and D. Di Stefano [7] that parity graphs are exactly those graphs that can be split decomposed into cliques and bipartite graphs.

References (35)

  • R. Anderson et al.

    Deterministic parallel list ranking

    Algorithmica

    (1991)
  • A. Barten, Design of Very Fast Parallel Algorithms in the Combinatorial Optimization, Diploma thesis, RWTH Aachen, 1989...
  • A. Bouchet

    Reducing prime graphs and recognizing circle graphs

    Combinatorica

    (1987)
  • F. Chin et al.

    Efficient parallel algorithms for some graph problems

    Comm. ACM

    (1982)
  • R. Cole, Parallel merge sort, inProceedings, 27th IEEE-FOCS, 1986, pp....
  • T. Cormen et al.

    Introduction into Algorithms

    (1990)
  • A. Cournier et al.

    A new linear algorithm for modular decomposition

  • Cited by (84)

    • Identifying clusters on a discrete periodic lattice via machine learning

      2019, Computer Physics Communications
      Citation Excerpt :

      All code can be found online at https://github.com/openerror/PhysicsLatticeClustering. There are many implementations of hierarchical clustering [13,14], some of which are parallelized [15–17]. In this paper we will work with the agglomerative variant, which forms clusters from the bottom-up — i.e. starting from single observations [4].

    • Marketing analytics: Methods, practice, implementation, and links to other fields

      2019, Expert Systems with Applications
      Citation Excerpt :

      For hierarchical clustering, Koga, Ishibashi, and Watanabe (2007) develop a method for agglomerative clustering that approximates the process of finding the nearest neighbor using hashing for choosing items to be merged, reducing the complexity of the algorithm. In addition, there have been several papers on how to best parallelize agglomerative clustering algorithms (Dahlhaus, 2000; Li, 1990; Olson, 1995). Similar methods exist for k-means clustering.

    View all citing articles on Scopus
    f1

    [email protected]

    1

    Present address: Institute of Computer Graphics, Vienna University of Technology.

    View full text