WEATHER PREDICTION USING MACHINE LEARNING TECHNIQUES

Weather forecasting plays a vital role in sectors such as agriculture, transportation, and disaster management. Traditional meteorological methods, like Numerical Weather Prediction (NWP) and statistical models, rely on physical and statistical formulations that are computationally intensive and limited by the chaotic nature of the atmosphere. In contrast, machine learning techniques such as Artificial Neural Networks (ANN) and Support Vector Machines (SVM) provide data-driven alternatives capable of learning complex relationships among atmospheric variables. This study proposes a hybrid ANN–SVM model for short- and medium-term weather forecasting, utilizing historical data including temperature, humidity, pressure, wind speed, and precipitation. The ANN component captures nonlinear dependencies among weather parameters, while the SVM ensures robustness against high-dimensional and noisy data. The system incorporates preprocessing, feature selection, and Explainable AI (XAI) techniques to improve interpretability and accuracy. Evaluation metrics such as RMSE and MAE are used to assess model performance. The hybrid approach demonstrates enhanced forecasting accuracy, robustness, and transparency, making it a promising alternative to conventional weather prediction systems.