Ανάλυση απώλειας και απόδοσης εργαζομένων : μια προσέγγιση με γνώμονα τα δεδομένα για τη διαχείριση ανθρώπινου δυναμικού
Analyzing employee attrition and performance : a data-driven approach for HR management
View/ Open
Keywords
Πρόβλεψη ; Μηχανική μάθηση ; Απώλεια εργαζομένων ; Ανάλυση δεδομένων ; Εφαρμογή πρόβλεψηςAbstract
This thesis, titled “Analyzing Employee Attrition and Performance: A Data-Driven Approach for HR Management,” studies the phenomenon of employee attrition, focusing on recent global trends such as The Great Resignation and burnout that drives employees to leave their jobs. The analysis focuses on the use of machine learning models—specifically, Random Forest, Logistic Regression, SVM, and XGBoost—to predict employee attrition. The process begins with Exploratory Data Analysis (EDA) to understand key employee data such as job satisfaction, total years of employment, and work-life balance, applying preprocessing and visualization to this data. Various models are trained on the given data and compared with performance parameters such as precision, accuracy, recall and F1 score. The best performing model is then used in a prediction application, where HR teams can input employee data and predict whether an employee is likely to leave the company. This prediction tool is a practical solution for improving employee retention strategies through proactive actions.