Кластеризация плазмид палочковидных форм бактерий и их видов с использованием спектроскопии

Abstract

This paper describes the solution of the problem of classification for bacterial species and their nutrient media. The described algorithm allows classifying three species of bacteria using self-organized map, as well as the clustering of three nutrient media for E. coli bacteria using k-means. The paper presents the rationale for choosing a solution model, calculating its accuracy, and ways to increase the accuracy. The experiments show the proposed solution is capable of clustering spectrograms with an accuracy of 96%.