Series Predictor : aplicativo para análise e previsão de séries temporais em Shiny

Abstract

Time series analysis is an essential tool across various domains of knowledge, enabling the understanding and prediction of temporal patterns in diverse phenomena. In this context, R emerges as a powerful platform for statistical analysis, offering extensive capabilities for data modeling and interpretation. However, its programmatic nature can pose a barrier to those who lack extensive programming experience. To overcome this challenge, this work introduces the development of an application based on R’s Shiny package, aimed at facilitating and democratizing time series analysis. With an intuitive and interactive interface, the application allows users to import data, transform them into time series, conduct trend and seasonality tests, adjust ARIMA/SARIMA models, and make future predictions. This initiative seeks to make time series modeling more accessible, enhancing informed decision-making across various sectors, from finance to health. By providing a user-friendly and effective tool, this study significantly contributes to the expansion of temporal analysis use, both in research and practice, promoting a deeper understanding of patterns and trends over time. Furthermore, the results observed through the use of the application demonstrate its effectiveness in providing valuable insights, consolidating its role as an indispensable tool in various fields of application.