In this article there is proposed a new two-parametrical variant of the gravitational classification method. We use the general idea of objects' behavior in a gravity field. Classification depends on a test object's motion in a gravity field of training points. To solve this motion problem, we use a simulation method. This classifier is compared to the 1NN method, because our method tends towards it for some parameter values. Experimental results on different data sets demonstrate an improvement in efficiency and that this approach outperforms the 1NN method by providing a significant reduction in the mean classification error rate.
@article{bwmeta1.element.doi-10_2478_bile-2014-0001,
author = {Tomasz G\'orecki and Maciej Luczak},
title = {A variant of gravitational classification},
journal = {Biometrical Letters},
volume = {51},
year = {2014},
pages = {1-12},
zbl = {06330995},
language = {en},
url = {http://dml.mathdoc.fr/item/bwmeta1.element.doi-10_2478_bile-2014-0001}
}
Tomasz Górecki; Maciej Luczak. A variant of gravitational classification. Biometrical Letters, Tome 51 (2014) pp. 1-12. http://gdmltest.u-ga.fr/item/bwmeta1.element.doi-10_2478_bile-2014-0001/