• data mining
• employee turnover
• explainability. HRPA
• human resources
• machine learning
• survey dataset

This thesis aims to predict employee turnover intention with machine learning and explainability techniques using a survey dataset. The thesis contains a review of the status of the art of HR turnover theories, a review of previous works using data mining and machine learning to predict employee turnover on survey data and explainability. After the literature review, the steps of the KDD process are put in place to perform the predictions. For the classification are used 5 algorithms: Logistic regression, random forest, decision tree, XGBoost, LightGBM. After the classification the results of the best performing algorithm are explained using Partial Dependence Plot.