Time series prediction based on machine learning: A case study, temperature forecasting in Vietnam

Abstract

In recent years, there has been a surge in interest in the subject of machine learning for prediction. In this study, a temperature dataset of Vietnam’s stations is examined in order to anticipate temperature. Several forecasting models are used to accomplish this goal. First, a traditional time-series forecasting approach such as Seasonal Autoregressive Integrated Moving Average is used (SARIMA). Then, more complex approaches such as XGBoost, Encoder-Decoder, and Prophet are used. The models' performances are compared using several accuracy assessment methods (e.g., Root Mean Squared Error (RMSE) and Mean Absolute Error (MAE)). The findings demonstrate the superiority of the deep learning approach over the other methods.