Prediction of Employee Turnover based on Machine Learning Models

Abstract

As a result of the fact that the procedure of decision making constitutes a vital component in the management of a company, the personnel of that company are seen as a valuable kind of asset by the latter. Therefore, the procedure of employing them in the first place by making the appropriate choices is generally recognised as a well-known obstacle by administrative authorities. Employee turnover may be a time-consuming and difficult process because recruiting new workers requires not only additional time but also a significant amount of financial expenditure. In addition to this, there are a number of additional elements that play a role in the selection and hiring of a qualified applicant, who in turn would provide economic returns for an organisation. In this research, I propose building a model to predict employee turnover rate using data from three datasets acquired from the Kaggle repository and a subset of their features. To analyse staff traits and forecast turnover and churn rate, the work summarised here employs machine learning approaches and pre-processing techniques. Logistic regression, AdaBoost, XGBoost, KNN, decision tress, and Naive Bayes are only some of the machine learning algorithms tried out on extracted datasets in the report's implementation experiments. Evaluating qualities against evaluation parameters like accuracy and precision factors follows thorough research and training of selected attributes.