Full text available in Adobe Acrobat format
Volume 51 Number 1 pp. 1-12
Tomasz Górecki 1, Maciej Luczak 2
1Faculty of Mathematics and Computer Science, Adam Mickiewicz University, Umultowska 87, 61-614Poznań, Poland
2Faculty of Civil Engineering, Environmental and Geodetic Sciences, Koszalin University of Technology, Śniadeckich 2, 75-453Koszalin, Poland
2Faculty of Civil Engineering, Environmental and Geodetic Sciences, Koszalin University of Technology, Śniadeckich 2, 75-453Koszalin, Poland
A variant of gravitational classification
<h2>Summary</h2><p> 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. </p>
Keywords: machine learning, nearest neighbor method, dynamic classifier, gravitational classification
DOI: 10.2478/bile-2014-0001
For citation:
| MLA | Górecki, Tomasz, and Maciej Luczak. "A variant of gravitational classification." Biometrical Letters 51.1 (2014): 1-12. DOI: 10.2478/bile-2014-0001 |
| APA | Górecki, T., & Luczak, M. (2014). A variant of gravitational classification. Biometrical Letters 51(1), 1-12 DOI: 10.2478/bile-2014-0001 |
| ISO 690 | GóRECKI, Tomasz, LUCZAK, Maciej. A variant of gravitational classification. Biometrical Letters, 2014, 51.1: 1-12. DOI: 10.2478/bile-2014-0001 |