APPLICATION OF THE METHOD OF K-NEAREST NEIGHBORS IN THE PROCESS OF INFORMATION AND ANALYTICAL SUPPORT OF ECONOMIC SECURITY OF DAIRY ENTERPRISES

Abstract

The article investigates the possibilities of classification of threats to dairy enterprises using the method of k-nearest neighbors. The purpose of this method is to solve problems of evaluating suppliers of raw materials for certain categories of reliability. Features of economic activity of milk processing enterprises are caused by constant search of high-quality raw materials. Of course, before using it, enterprises use a number of inspections, but such inspections are quite expensive and are not always carried out for various reasons. The method of information classification studied in the work can be used with the use of machine learning technologies. For its approbation the program code which includes such libraries as math, random, pylab, numpy, matplotlib was used. To classify suppliers of raw materials, a test sample of data was created, after which the distance to each of the objects of the training sample was calculated. Selected k objects of the training sample, the distance to which is minimal. And also the class of the supplier of raw materials which meets criteria of reliability is defined. Part of the attention is paid to the characteristics of suppliers of raw materials for dairy enterprises: conditions for keeping cows, approaches to the collection of raw materials, storage of raw materials and their transportation. Based on the study of the peculiarities of the suppliers of raw materials for dairy enterprises, quantitative criteria for their reliability have been established. Thus, the relevant criteria were evaluated using the method of k-nearest neighbors, which allowed to form a classification of suppliers of raw materials. At the same time, the article highlights the characteristics of information and analytical support of economic security of dairy enterprises, highlights a number of key features of such activities.