Инновации. Наука. Образование Роза Виталий Петрович Роза Мария Петровна Студент магистратуры и аспирантка
Сибирский Государственный Университет
Имени академика М. Ф. Решетнева
METHODS OF CLUSTER DATA ANALYSIS AND CLUSTERING QUALITY ASSESSMENT Аннотация: В данной статье определяется важность использования методов
автоматической группировки, а также приводится оценка качества кластеризации.
Ключевые слова: анализ данных, кластеризация, автоматическая группировка, кластеры. Keywords: data analysis, clustering, automatic grouping, clusters. Modern Russian researchers present automatic grouping problems in the form of integer
linear programming problems. Currently, there are a huge number of different effective methods
for solving optimization problems. It happens that difficulties arise during the solution of large
tasks of automatic grouping, taking into account the fact that there is an instantaneous increase in
the volume of data that is collected and processed in automated systems. Clustering, based on the
established similarity relation of elements, establishes subsets (clusters) into which the input data
is grouped. One of the simplest and most effective are methods and models based on minimizing
the total distances between objects of the same group (cluster) or between cluster objects and its
center.
Automatic grouping methods are used in many branches of science, including actively
used in data mining systems. Due to the use of models of optimal placement and automatic
grouping of objects, the requirements for economic efficiency are increased. Automatic grouping
methods can group objects by constructing models of the relationship of objects in a continuous
space of characteristics. Such methods have the opportunity to be applicable with sufficiently
large amounts of data. At the same time, tasks should be solved interactively with a limited
working time with a large amount of input data.
816
Научный журнал «Инновации. Наука. Образование» Индексация в РИНЦ н