Séries temporais como ferramenta de previsão de preços pagos ao produtor de cana-de-açucar no Nordeste (AL, PB e PE)

Abstract

The cultivation of sugarcane is a key activity in Northeast Brazil, serving as a significant economic source for its states and generating a considerable number of jobs in the region. Its importance extends further when considering the numerous possibilities regarding its derived products and by-products. Consequently, decision-making by stakeholders in the production chain becomes crucial for maintaining this source of employment and income, particularly with regard to prices. In light of this, the present research explored the use of the Python programming language and applied time series techniques for price forecasting in a specific context, focusing on sugarcane cultivation in the states of Alagoas, Pernambuco, and Paraíba from 2014 to 2022. Throughout this study, it was identified that the AR(1) model would be a simple yet efficient approach to achieve the expected results. In the modeling process, this model was implemented, and the results obtained demonstrate the feasibility and effectiveness of utilizing time series and autoregressive models for price forecasting in the studied context. However, further research is necessary to determine whether this model is the most suitable and accurate for prediction, as a comparative analysis of results specific to the circumstances of the study is ideal.