Predicting employee work durations using machine learning algorithms

Employee turnover rate may vary depending on many factors such as the structure of the sector, wage policy, labor market conditions, quality of life offered by cities and alternative job opportunities. High turnover rate can lead to negative consequences such as increased cost, loss of productivity, decrease in motivation and decrease in customer satisfaction. Therefore, understanding and managing employee turnover is critical to the efficiency and sustainable success of businesses. In this study, it is aimed to more accurately estimate the length of stay of employees in the enterprise by considering the factors affecting the employee turnover rate with a multidimensional approach. For this purpose, data of employees who left their jobs in a company in the last four years were used and these data were examined with both statistical and machine learning approaches. Running times were estimated using K-Nearest Neighbor (KNN) and Naive Bayes (NB) methods, which are machine learning algorithms. This study demonstrates the applicability of machine learning methods to effectively predict employee turnover and provides a significant contribution to businesses in making data-driven, informed decisions in their human resources strategies.