Banca de QUALIFICAÇÃO: EDUARDO JACOMO SERAPHIM NOGUEIRA

Uma banca de QUALIFICAÇÃO de MESTRADO foi cadastrada pelo programa.
STUDENT : EDUARDO JACOMO SERAPHIM NOGUEIRA
DATE: 23/12/2024
TIME: 14:00
LOCAL: Vídeo Conferência
TITLE:

Hybrid Modeling Fiscal Time Series: Integration of Econometric and Machine Learning Models for Macroeconomic Forecasting in Brazil.

KEY WORDS:

Fiscal Forecasting; Time Series; Hybrid Modeling.

PAGES: 20
BIG AREA: Ciências Sociais Aplicadas
AREA: Economia
SUMMARY:

This project aims to investigate the application of models for forecasting fiscal time series by integrating traditional econometric approaches with advanced machine learning techniques. The research seeks to improve the accuracy of fiscal projections through the combination of models such as ARIMA and VAR with LSTM neural networks and random forests. Integration strategies and comparative performance analysis are explored to capture nonlinear patterns and complex seasonality in Brazilian fiscal variables. The approach aims to provide practical contributions to institutions that monitor fiscal policy implementation in the country, such as the Independent Fiscal Institution (IFI), by proposing more robust methods for fiscal monitoring and public policy formulation based on more accurate and reliable forecasts.

COMMITTEE MEMBERS:
Externo à Instituição - JOÃO FROIS CALDEIRA - UFSC
Presidente - 1642911 - DANIEL OLIVEIRA CAJUEIRO