Abstract:

The forecast of the degradation of employees in the work life balance is an important indicator for understanding the performance of employees for the human resources department in companies. For such indicators, predicting with data mining instead of classical statistical analyses will give more effective results. In this study, the estimated situation of degradation in the business life and therefore leaving the work was compared with the performance results by making decisions from the data mining methods with the methods of the tree and support vector machine. The sample data set for analysis was taken from a database provided under the IBM Watson Analytics program. The data set contains 1470 employees’ 35 different subjects. The study identified the factors affecting the breakdown and the estimated performance of the decision tree and support vector machine methods was achieved at 84,09% and 91,36% accuracy rates, respectively.

Abstract:

The estimation of employee attrition in business life balance is an important indicator to understand the employee performans for human resoırces department in companies. For such indicators, rather than classical statistical analysis, estimating with data mining will provide more effective results. In this study, the prediction of attrition and hence the turnover situations were estimated with decision tree and support vector machine methods from data mining and performace result were compared. The sample dataset for analysis was taken from a database provided by the IBM company in scope of Watson Analytics programme. The data set contains 35 different attributes of 1470 employees. Factors affecting attrition were determined in the study and the estimation performance of the decision tree and support vector machine methods were obtained for accuracy rates of 84.09% and 91.36% respectively.

Keywords: