The problem of the formation of generalized concepts when working with uncertain attributes

I. I. Astakhova, V. N. Vagin, M. V. Fomina

National Research University "Moscow Power Engineering Institute", Moscow

Abstract: This paper considers the problem of data generalization, taking into account uncertain information in attributes. Algorithms capable of operating with uncertain data are considered. We suggest methods that improve UDT algorithm. The results of experiments on the application of the proposed algorithms are given.

Keywords: generalization problem, undefined numeric attribute, decision tree.

UDC: 621.398

Received: 04.09.2017
Revised: 15.12.2017

DOI: 10.26456