A variant of gravitational classification

• Autorzy: Tomasz Górecki , Maciej Luczak
• Języki publikacji: EN

Abstrakty

EN

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.

Słowa kluczowe

EN

machine learning; nearest neighbor method; dynamic classifier; gravitational classification

• Wydawca: De Gruyter Open
• Czasopismo: Biometrical Letters
• Rocznik: 2014
• Tom: 51
• Numer: 1
• Strony: 1-12

Daty

wydano

2014-06-01

online

2014-06-06

Twórcy

autor

Tomasz Górecki

• Faculty of Mathematics and Computer Science, Adam Mickiewicz University, Umultowska 87, 61-614 Poznań, Poland

autor

Maciej Luczak

• Faculty of Civil Engineering, Environmental and Geodetic Sciences, Koszalin University of Technology, Śniadeckich 2, 75-453 Koszalin, Poland

Identyfikatory

DOI

10.2478/bile-2014-0001