Redes neuronales recurrentes y modelos arima para el pronóstico de la inflación en el Perú

• Autor: Chilón Ayay, Anghie Lizzeth
• Formato: tesis de grado
• Fecha de Publicación: 2023
• Institución: Universidad Nacional de Trujillo
• Repositorio: UNITRU-Tesis
• Lenguaje: español
• OAI Identifier: oai:dspace.unitru.edu.pe:20.500.14414/18318
• Enlace del recurso: https://hdl.handle.net/20.500.14414/18318
• Nivel de acceso: acceso abierto
• Materia: Modelos ARIMA; Redes Neuronales Recurrentes; LSTM; Inflación

The purpose of this research is to compare the ARIMA methodology and Recurrent _x000D_ Neural Networks to find the best model to forecast Inflation in Peru. The research consisted _x000D_ of a study sample from January 2000 to December 2021._x000D_ When designing the series model for each series, a forecast was made for all the _x000D_ months of the year 2022, which were compared with the real data to determine which _x000D_ methodology makes a better forecast. The results indicated that the previous LSTM recurrent _x000D_ neural network model obtained a lower forecast evaluation error compared to the ARIMA _x000D_ methodology, for which the RMSE, MAE, MAPE, EMC and MPE indicators were used