Employee attrition prediction using machine learning models

• Autores: Iparraguirre-Villanueva, Orlando, Chauca-Huete, Luis, Paulino-Moreno, Cleoge
• Formato: objeto de conferencia
• Fecha de Publicación: 2024
• Institución: Universidad Tecnológica del Perú
• Repositorio: UTP-Institucional
• Lenguaje: inglés
• OAI Identifier: oai:repositorio.utp.edu.pe:20.500.12867/14263
• Enlace del recurso: https://hdl.handle.net/20.500.12867/14263; https://doi.org/10.18687/LACCEI2024.1.1.498
• Nivel de acceso: acceso abierto
• Materia: Machine learning; Artificial intelligence; Management; Human Resources; https://purl.org/pe-repo/ocde/ford#2.02.04

Today's business landscape is characterized by competition and dynamism, which has transformed human resource management into an essential strategic partner for organizations. Employee turnover poses risks that affect productivity and knowledge management. This study focuses on predicting employee turnover using machine learning (ML) models. For the training process, a dataset composed of 4410 records and 29 variables was used, in the process of training and evaluation of the ten models, the artificial intelligence (AI) method was followed. The findings showed that the XG Boost Classifier (XGBC) and Random Forest (RF) models achieved the best accuracy and performance rates, with 98.8% and 98.7%. Followed by Decision Tree Classifier (DT) with 97.6%, and the other models, such as Gradient Boosting Classifier (GBC), Ada boost Classifier (AC), Logistic Regression (LR), KN Classifier (K-NNC), SGD Classifier (SGDC), Support Vector Classifier (SVC) and Nu Support Vector Classifier (NuSVC), achieved the following rates: 88.4%, 85.4%, 84%, 82.2%, 83.0%, 83.0%, 55.0%, respectively. Finally, it is concluded that the models are useful and effective in prediction. Their practical implementation in human resource management strategies is recommended for proactive intervention.