Data Clustering

Data Clustering

Bülent Başaran (Bilecik Şeyh Edebali University, Turkey) and Fatih Güneş (Bilecik Şeyh Edebali University, Turkey)
Copyright: © 2017 |Pages: 45
DOI: 10.4018/978-1-5225-1776-4.ch002
OnDemand PDF Download:
$30.00
List Price: $37.50

Abstract

Data clustering can be associated with and used in many research methodologies and application areas. In this chapter, the basics of data clustering and some kind of its applications are given with examples and a real data set. The examples show data types and help to explain basic clustering algorithms. The real data help to classify countries in terms of their computer and internet proficiency levels. After examining the resulting clusters obtained from four different basic methods according to eight computer and internet proficiencies, it is found that the regional closeness of countries and being outside of a union are the major drivers of those clusters' formation. This application gives some possible future research area extensions to researchers about clustering the same or other countries by different proficiencies and unions.
Chapter Preview
Top

Application Areas Of Data Clustering

Because it is the most widely used technique for grouping data, and separation of the data into groups according to some attributes is necessary in many processes, procedures, disciplines, and strategic issues, data clustering has many application areas spread all over the life. Those areas include pattern recognition, data analysis, image processing, outlier detection, market segmentation, facilities layout, natural sciences etc. Aggarwal and Reddy (2014) approach those issues from the application domains in which the clustering problem arising point of view. They listed the problem areas as intermediate step for other fundamental data mining problems, collaborative filtering, customer segmentation, data summarization, dynamic trend detection, multimedia data analysis, biological data analysis, and social network analysis. DeSarbo and Mahajan (1984) approach them from external and/or internal information (reasons) on the objects to be classified. External reasons do not depend on the physical resemblances of the objects whereas internal reasons depend on those physical resemblances of the objects. Punj and Stewart (1983) present marketing research application areas. Those areas are market segmentation, identifying homogeneous groups of buyers, development of potential new product opportunities, test market selection, and general data reduction. Gong and Richman (1995) give the application areas of atmospheric science research. Wilk and Pelka (2013) give numerical data application areas such as classification of people, plants, animals and chemical elements, and symbolic data application areas such as data mining, artificial intelligence, machine learning research, cognitive psychology research and conceptual clustering.

Complete Chapter List

Search this Book:
Reset