Abstract
The k-Nearest Neighbors Graph (kNNG) consists of links from an object to its k-Nearest Neighbors. This graph is of interest in diverse applications ranging from statistics, machine learning, clustering and outlier detection, computational biology, and even indexing. Obtaining the kNNG is challenging because intrinsically high dimensional spaces are known to be unindexable, even in the approximate case. The cost of building an index is not well amortized over just all the objects in the database. In this paper, we introduce a method to compute the kNNG without building an index. While our approach is sequential, we show experimental evidence that the number of distance computations is a fraction of the \(n^2/2\) used in the naïve solution. We make heavy use of the notion of pivot, that is, database objects with full distance knowledge to all other database objects. From a group of pivots, it is possible to infer upper bounds of distance to other objects.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
References
Archip, N., Rohling, R., Cooperberg, P., Tahmasebpour, H., Warfield, S.K.: Spectral clustering algorithms for ultrasound image segmentation. In: Duncan, J.S., Gerig, G. (eds.) MICCAI 2005. LNCS, vol. 3750, pp. 862–869. Springer, Heidelberg (2005). https://doi.org/10.1007/11566489_106
Baeza-Yates, R., Hurtado, C., Mendoza, M.: Query clustering for boosting web page ranking. In: Favela, J., Menasalvas, E., Chávez, E. (eds.) AWIC 2004. LNCS (LNAI), vol. 3034, pp. 164–175. Springer, Heidelberg (2004). https://doi.org/10.1007/978-3-540-24681-7_19
Brito, M., Chávez, E., Quiroz, A., Yukich, J.: Connectivity of the mutual k-nearest neighbor graph in clustering and outlier detection. Stat. Probab. Lett. 35(4), 33–42 (1996)
Callahan, P., Kosaraju, R.: A decomposition of multidimensional point sets with applications to k nearest neighbors and n body potential fields. JACM 42(1), 67–90 (1995)
Chávez, E., Ludueña, V., Reyes, N.: Solving all-k-nearest neighbor problem without an index. In: Procs. del XXV Congreso Argentino de Ciencias de la Computación (CACIC 2019), pp. 567–576. UniRío editora (2019)
Chávez, E., Ludueña, V., Reyes, N., Roggero, P.: Faster proximity searching with the distal sat. Inf. Syst. 59, 15–47 (2016)
Chávez, E., Navarro, G.: A compact space decomposition for effective metric indexing. Pattern Recogn. Lett. 26(9), 1363–1376 (2005)
Chávez, E., Navarro, G., Baeza-Yates, R., Marroquín, J.: Searching in metric spaces. ACM Comput. Surv. 33(3), 273–321 (2001)
Duda, R., Hart, P.: Pattern Classification and Scene Analysis. Wiley, New York (1973)
Eppstein, D., Erickson, J.: Iterated nearest neighbors and finding minimal polytopes. Int. J. Math. Comput. Sci. 11–3, 321–350 (1994)
Figueroa, K., Navarro, G., Chávez, E.: Metric spaces library (2007). http://www.sisap.org/Metric_Space_Library.html
Navarro, G., Paredes, R., Reyes, N., Bustos, C.: An empirical evaluation of intrinsic dimension estimators. Inf. Syst. 64, 206–218 (2017)
Paredes, R.: Graphs for Metric Space Searching. Ph.D. thesis, University of Chile, Chile, July 2008
Paredes, R., Chávez, E., Figueroa, K., Navarro, G.: Practical construction of k-nearest neighbor graphs in metric spaces. In: Àlvarez, C., Serna, M. (eds.) WEA 2006. LNCS, vol. 4007, pp. 85–97. Springer, Heidelberg (2006). https://doi.org/10.1007/11764298_8
Rubinstein, A.: Hardness of approximate nearest neighbor search. In: Proceedings of the 50th Annual ACM SIGACT Symposium on Theory of Computing, pp. 1260–1268. ACM (2018)
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2020 Springer Nature Switzerland AG
About this paper
Cite this paper
Chávez, E., Ludueña, V., Reyes, N. (2020). Heuristics for Computing k-Nearest Neighbors Graphs. In: Pesado, P., Arroyo, M. (eds) Computer Science – CACIC 2019. CACIC 2019. Communications in Computer and Information Science, vol 1184. Springer, Cham. https://doi.org/10.1007/978-3-030-48325-8_16
Download citation
DOI: https://doi.org/10.1007/978-3-030-48325-8_16
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-030-48324-1
Online ISBN: 978-3-030-48325-8
eBook Packages: Computer ScienceComputer Science (R0)