ENTROPİ-CRITIC TEMELLİ EM ALGORİTMASI İLE İNTERNET KULLANIMI AÇISINDAN OECD ÜLKELERİNİN KÜMELENMESİ

In the study, mobile and cable internet subscription rates were determined as two different criteria and the clustering of 37 OECD countries between 2012 and 2019 was made. The weights of the criteria were respectively calculated using the Entropy method and CRITIC method. The weights are reflected in data created for the groups of 37 OECD countries divided into 2012-2015 and 2016-2019, making a total of 4 different clustering operations. By determining the similarities and differences in Internet subscription rates of OECD countries by year, changes between clusters were determined. Finally, the Wilcoxon test was performed to determine the significance levels of differences between internet uses, and the Kruskal Wallis-H test was performed to analyze the differences between 5 clusters formed according to the EM algorithm. It was concluded that the increase in internet use in the period 2016-2019 compared to internet use in the period 2012-2015 was statistically significant.