Comparison Of Iron And Steel Production Defects Using Classification Algorithms

The iron and steel industry are a strong foundation of economic development in the world. The iron and steel industry are directly related to the economic developments in the world market and the economic powers of the countries. It also provides input to all branches of industry. Data mining and techniques in this sector play an important role in the activities of corporate and large enterprises as a scientific method. It involves the processes of finding and modeling meaningful relationships between meaningless large chunks of data in an enterprise. Studies on the iron and steel industry in the world and in our country are very limited. Analyzing products in the iron and steel industry using data mining techniques will both save time and contribute to reducing the financial burden of operators. It is also thought that it can increase the preferability level by increasing the quality of the products offered to the customers. In this study, the data mining process of the iron and steel industry is defined and the data mining studies applied to some quality improvement problems in the production sector are examined, and the optimization of the process and quality parameters from the quality improvement problems are emphasized. In the application section, data mining techniques are used to determine the variables and levels that cause manufacturing defects in an industrial enterprise. To achieve this goal, the decision tree, which is one of the data mining classification methods, has been applied with the C5.0, C&RT, CHAID and QUEST algorithms, the decision tree has been created with the highest accuracy C5.0 algorithm and the results have been examined. As a result of the analysis, the products produced by the industrial enterprise are classified according to production defects.