Prediction with machine learning methods – an application inthe iron and steel industry

Firms aim to have a low turnover rate since the cost of retaining existing staff is lower than the cost of hiring a new employee. Considering that the iron and steel industry has a high turnover rate due to the tough working conditions. In this study, machine learning algorithms were used and it was aimed to predict the personnel who are likely to leave the job. Moreover, machine learning algorithms were implemented using Knime and Python programming language. The dataset consists of 2318 rows, each corresponding to one employee, and 14 different attributes over a period of approximately 16 years. The attributes were selected from the attributes that are thought to have an impact on the resignation of employees. Five different machine learning algorithms were applied, and it was observed that the Random Forest Algorithm provide most successful results with an accuracy rate of 78.40% in Knime and 76.88% in Python programming language. According to the Random Forest Algorithm, age and seniority are the most important criteria by ranking the most important qualifications for leaving a job. With the Random Forest Algorithm, the attributes were valued by making assumptions and it was evaluated whether the considered personnel had the potential to leave the job. As a result of this study, it was determined whether the existing staff has the potential to leave the job, which criteria should be prioritized in recruitment, and which measures should be taken to ensure the continuity of the staff.