Human resources analytics in recruitment practices

It has been seen recently that the analytical studies are extremely critical for the development of a data-based decision-making culture in Human Resources Management, both in strategic human resources decisions and operational processes in the digitalization journey. With this thesis, the function of selecting and placing the employees, which is one of the most important functions of human resources management, was examined, its relationship with HR analytics was mentioned and analyses were made by establishing a modelling. In addition to these, a real-time estimation of the employee needs was provided. In line with this purpose, a data set containing the attributes of the candidates whose job interview was held and whose outcome (target variable) was certain, was used. There were 266 people in this data set. The obtained data were evaluated by applying logistic regression, random forest, gradient boosting decision tree, naïve bayes analysis over the R program. At the next stage, in a new data set consisting 11 people whose outcomes were uncertain, estimations were made with the random forest method, which had the highest accuracy rate of 89% among these analyses. As a result of these estimations, 2 people with the positive results and 9 people with the negative results were obtained.