Comparison of the Accuracy of Linear Regression and Support Vector Regression in Average Temperature Prediction

Abstract

The weather in Indonesia varies significantly and is influenced by geographical location, topography, and regional climate. Weather patterns differ between the western and eastern parts of Indonesia. This study explores time series models to predict weather data in Palu City, a region that is complex due to various weather factors. The focus is on the unique weather patterns reflected by the geography and topography of Palu City. Evaluation was conducted on time series models, including Linear Regression and Support Vector Regression (SVR), to estimate weather conditions in Palu City. The evaluation results show that the SVR model has an RMSE of 0.6302, while linear regression has an RMSE of 0.6328. This research has the potential to improve early warning and decision-making regarding extreme weather