Techniques of machine learning applied to reduce employee turnover in a company cleaning and disinfection

• Autor: Romero Rojas, Erika Noemi
• Formato: tesis de grado
• Fecha de Publicación: 2024
• Institución: Universidad de Lima
• Repositorio: ULIMA-Institucional
• Lenguaje: inglés
• OAI Identifier: oai:repositorio.ulima.edu.pe:20.500.12724/20339
• Enlace del recurso: https://hdl.handle.net/20.500.12724/20339
• Nivel de acceso: acceso abierto
• Materia: https://purl.org/pe-repo/ocde/ford#2.11.04

Staff turnover in large Peruvian manufacturing industries has been increasing in recent years. While job rotation is a natural effect in organizations, it generates higher training costs for new staff and impacts work performance and climate when unwanted. Given this problem arises the need to identify the possible causes of rotation of operational personnel and predict these events through data analysis at an early stage to avoid and reduce its impact on the company. This article of quantitative approach and exploratory scope-explanatory aims to identify the propensity of rotation of the operation of a company manufacturing cleaning and disinfection through a model of forecast by collecting data using Machine Learning and encourage proposals that enable solutions to be found to the factors influencing staff turnover. MS Excel and Orange software were used for data analysis, where the data were trained with different intelligence models such as Random Forest, Logistic Regression, Decision Tree, and SVM, and Python to run the model and get numerical indicators like the Area under the curve (AUC) and the analysis of the ROC curve. The proposed study shows that the models perform well in classification, with high accuracy and recall rates, 96% and 97%, respectively, and an overall accuracy of 96%.