An Ensemble Approach for Improving Time-series Weather Data Accuracy

Abstract

The weather climate is contingent upon various meteorological factors, including humidity, temperature, and pressure, among others. The sudden alteration of these factors within the surrounding atmosphere begets detrimental consequences, spanning from the production sector to healthcare. As a result, weather prediction is required to prevent potentially hazardous situations with minimum loss. This paper presents an ensemble model that enhances the accuracy of predicting time-series weather data. The model has been examined using datasets from two different cities, namely Bengaluru and Dongsi. It has been observed that the model demonstrates superior accuracy in forecasting temperature fluctuations, seasonal patterns, and long-term climate variations. Furthermore, the model consistently outperforms several stateof-the-art models in terms of performance metrics such as MAE, MSE, RMSE, and R 2 Score.