Improving Demand Analysis and Supply Chain Management for Hair Products During the COVID-19 Pandemic: A Machine Learning Approach

• Autores: Kato Yoshida, Valeria Midori, Mosquera Mendoza, Ivone Brigiethe, García López, Yván Jesús, Quiroz Flores, Juan Carlos
• Formato: artículo
• Fecha de Publicación: 2023
• Institución: Universidad de Lima
• Repositorio: ULIMA-Institucional
• Lenguaje: inglés
• OAI Identifier: oai:repositorio.ulima.edu.pe:20.500.12724/19479
• Enlace del recurso: https://hdl.handle.net/20.500.12724/19479; https://doi.org/10.14445/22315381/IJETT-V71I9P234
• Nivel de acceso: acceso abierto
• Materia: Supply chain; Hair preparations; Machine learning; Pandemics; Cadena de suministro; Productos capilares; Aprendizaje automático; Pandemias; COVID-19; https://purl.org/pe-repo/ocde/ford#2.02.03

This research analyzes the demand for hair care products during the COVID-19 pandemic. Two forecasting models, Arima and Sarima, based on Machine Learning technology, were proposed to improve data analysis and supply chain management. The results showed that the SARIMA model had higher mean absolute error levels than the Arima model. The study also analyzed the demand for four hair dyes using statistical models, finding that three had seasonal demand. The SARIMA model accurately predicted demand for most hair dyes except one. Errors in the predictions were measured using different indicators, and the SARIMA model had lower error levels than the Arima model. The study's results were validated and compared with previous research, showing that the SARIMA model predicted the demand for hair dyes. Overall, this study highlights the usefulness of Machine Learning models in demand analysis and supply chain management of hair care products during the COVID-19 pandemic. These findings provide a reference framework for manufacturing industries with similar characteristics that wish to optimize demand management using Machine Learning techniques.