Previous |  Up |  Next


nearest neighbour classification; prototype selection
Prototype Selection (PS) techniques have traditionally been applied prior to Nearest Neighbour (NN) classification rules both to improve its accuracy (editing) and to alleviate its computational burden (condensing). Methods based on selecting/discarding prototypes and methods based on adapting prototypes have been separately introduced to deal with this problem. Different approaches to this problem are considered in this paper and their main advantages and drawbacks are pointed out along with some suggestions for their joint application in some cases.
[1] Devijver P. A., Kittler J.: Pattern Recognition. A Statistical Approach. Prentice Hall, 1982 MR 0692767 | Zbl 0624.00015
[2] Ferri F., Vidal E.: Colour image segmentation and labelling through multiedit–condensing. Pattern Recognition Lett. 13 (1992), 8, 561–568
[4] Gates G. W.: The reduced nearest neighbor rule. IEEE Trans. Inform. Theory 18 (1972), 431–433 DOI 10.1109/TIT.1972.1054809
[5] Geva S., Sitte J.: Adaptive nearest neighbour pattern classification. IEEE Trans. Neural Networks 2 (1991), 2, 318–322 DOI 10.1109/72.80344
[6] Hart P. E.: The condensed nearest neighbor rule. IEEE Trans. Inform. Theory 14 (1968), 515–516 DOI 10.1109/TIT.1968.1054155
[7] Kohonen T.: Self–Organization and Associative Memory. Second edition. 1988 MR 0930103 | Zbl 0659.68100
[8] Kohonen T., Kangas J., Laaksonen J., Torkkola K.: Lvq_pak: The Learning Vector Quantization Program Package. Technical Report, Helsinki Univ. of Tech., 1992
[9] Kraaijveld M. A., Duin R. P. W.: On backpropagation learning of edited data sets. In: Proc. of the Int. Neural Network Conf., 1990, pp. 741–744
[10] Lucas A. E., Kittler J.: A comparative study of the kohonen and multiedit neural net learning algorithms. In: 1st IEE Int. Conf. Artificial Neural Networks, 1991, pp. 7–11
[11] Wilson D. L.: Asymptotic properties of nearest neighbor rules using edited data. IEEE Trans. Systems Man Cybernet. 2 (1972), 3, 408–421 DOI 10.1109/TSMC.1972.4309137 | MR 0329139 | Zbl 0276.62060
Partner of
EuDML logo