An Algorithm For Selecting Informative Symbols Based On Determining The Measure Of Information In A Nominally Informed Space

The article reveals the issue of reducing the size of the phase of features describing objects, from data mining to brain cancer diseases. Initially, with the support of specialists in the field of medicine, 218 objects of the 4th class were conducted (it is a paid astrocytoma of the right hemisphere of the brain; Adenoma of the cellar region of the brain; Glioblastoma of the right-sided region of the brain; Meningioma of the right frontal region of the brain) and a training sample of 19 characters is formed. In this educational selection, the features characterizing the objects of the class are expressed as a nominal value. For this reason, this article proposes an algorithm for solving the problem of choosing a set of informative symbols based on determining the measurement of information in a nominal data space.

Hali tarjima qilinmagan