Clustering of OECD countries' foreign trade data with SOM algorithm

Countries have different development, prosperity, economic and commercial capacities in the international arena. Each country aims to reach the highest level and develops its own strategies for this. They realize these strategies by following global conditions and standards. The sample countries selected in this study were examined in accordance with 13 pre-agreed criteria. In this research, which consists of three parts, the history of data mining developed with technology is briefly explained in the first part, and then a detailed explanation is given about the clustering methods, which is the method of the study. In the second part, subjects such as the history, organizational structure and membership conditions of the Organization for Economic Cooperation and Development (OECD), which was established in 1948 under the leadership of the Marshall Plan aiming to develop the European countries devastated after the Second World War, and became influential globally with the participation of the United States and Canada in 1960, has been studied in depth. In the last part, the implementation part, clusters were obtained on the decision matrix consisting of countries and criteria. These criteria are 13 in total, and they are inflation, foreign investment, imports, unemployment, population, gross domestic product, per capita income, purchasing power parity, exports, employment rates by age group (15-24, 25-54, 55-64), and tax statistics. After obtaining the decision matrix consisting of countries and determined criteria, 2019 data from the World Bank and OECD were included in the matrix. Then, the countries were divided into clusters with the Weka Program and the Kruskal-Wallis Test was applied to find the significant difference of the divided clusters, and finally Dunn Index was used to calculate the optimal value of the data.