Abstract
In a finite set X with distance, we introduce a so-called chain distance. This distance generates a partition of X into clusters such that any point inside each cluster can be connected with any other point of the same cluster by a chain whose every link does not exceed a given threshold value. We construct a chain development, by which we mean a mapping of X into a straight line that preserves the chain distance and allows one to rapidly perform clustering. We also present an efficient algorithm for constructing a chain development.
Similar content being viewed by others
References
A. Yu. Lemin, “Isometric imbedding of isosceles (non-Archimedean) spaces in Euclidean spaces,” Dokl. Akad. Nauk SSSR 285 (3), 558–562 (1985)
A. Yu. Lemin, Sov. Math., Dokl. 32, 740–744 (1985)].
R. C. Prim, “Shortest connection networks and some generalizations,” Bell Syst. Tech. J. 36, 1389–1401 (1957).
V. Klyahzkin, E. Shchepin, and K. Zingerman, “Application of hierarchical methods of cluster analysis to the printed text structure recognition,” in Shape, Structure and Pattern Recognition: Nahariya (Israel), October 4–6, 1994 (World Sci., Singapore, 1995), pp. 333–342.
V. M. Klyatskin, E. V. Shchepin, and K. M. Zingerman, “Hierarchical analysis of multicolumn texts,” Pattern Recognit. Image Anal. 5 (1), 87–98 (1995).
N. V. Kotovich, “Algorithms for clustering images of symbols,” Tr. Inst. Sist. Anal., Ross. Akad. Nauk 38, 241–251 (2008).
Author information
Authors and Affiliations
Corresponding author
Additional information
Original Russian Text © Yu.V. Malykhin, E.V. Shchepin, 2015, published in Trudy Matematicheskogo Instituta imeni V.A. Steklova, 2015, Vol. 290, pp. 317–322.
Rights and permissions
About this article
Cite this article
Malykhin, Y.V., Shchepin, E.V. Chain development. Proc. Steklov Inst. Math. 290, 300–305 (2015). https://doi.org/10.1134/S0081543815060267
Received:
Published:
Issue Date:
DOI: https://doi.org/10.1134/S0081543815060267