Abstract
In the best conditions, the time complexity of classical algorithms of hierarchical ascendent classification is for a number n of elements to be classified of order O(n2). In this paper the proposed CAHCVR algorithm is based on the principle of the “reciprocal neighbourhood” algorithm (CAHVR). It takes into account a contiguity constraint defined by a given connected graph. We study the properties of its results (absorbing class, inversion,…) and demonstrate that the average time complexity is in O(n) (linear complexity). The proposed CAHCVR algorithm is applied to image analysis in order to solve the image segmentation problem.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Preview
Unable to display preview. Download preview PDF.
References
Bachar, K [1994] Contributions en analyse factorielle et en classification ascendante hiérarchique sous contrainte de contiguïté. Applications à la segmentation d’images. Thèse de doctorat, Université de Rennes 1 (France).
Benzécri, J-P [1982] Sur la proportion des paires de voisins réciproques pour une distribution uniforme dans un espace euclidien. Les cahiers de l’analyse des données, Vol. 7, p. 185–188.
Lebart, L [1978] Programme d’agrégation avec contraintes. Les cahiers de l’analyse des données, Vol. 3, p. 275–287.
Lerman, I-C [1991] Foundations of the Likelihood Linkage Analysis(LLA) Classification method. Applied Stochastic Models and Data Analysis, Vol.7, #1, march 1991, p.63–76, John Wiley.
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 1998 Springer-Verlag Berlin · Heidelberg
About this paper
Cite this paper
Bachar, K., Lerman, IC. (1998). Statistical Conditions for a Linear Complexity for an Algorithm of Hierarchical Classification Under Constraint of Contiguity. In: Rizzi, A., Vichi, M., Bock, HH. (eds) Advances in Data Science and Classification. Studies in Classification, Data Analysis, and Knowledge Organization. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-642-72253-0_18
Download citation
DOI: https://doi.org/10.1007/978-3-642-72253-0_18
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-540-64641-9
Online ISBN: 978-3-642-72253-0
eBook Packages: Springer Book Archive