Comparison of hierarchical cluster methods by cophenetic correlation in big data

In this study, firstly, theoretical information about the definition of big data, components of big data, Big data analytics and big data technologies are included. In addition, theoretical information about cluster analysis, clustering methods, distance measures of clustering method and cophenetic correlation coefficient are given. Afterwards, hierarchical clustering methods in big data using big data technologies were compared with the cophenetic correlation coefficient. Amazon Cloud Server containing open source big data technologies was used for data analysis. Python programming language is installed on this server. Libraries developed for Python were used in the analysis processes. Air Travel Consumer Report in the USA for 2015, which was published as an open access data set, was used. Since the inclusion of variables that do not affect the result analysis may prolong the analysis process, the feature selection process has been performed. The blank observations were then cleared and the data were standardized. Afterwards, 4 different data sets were created by random selection method representing the main population from the data set. Clustering analysis was applied to these data sets. As a result of the analysis, it was observed that the cophenetic correlation coefficient gave the highest result in the Avarage Clustering method in all data sets. 2020, ix + 50 pages Keywords: Cophenetic correlation, Big data, Cluster analysis.