Machine Learning for Temperature Forecasting in Bangladesh
Machine Learning for Temperature Forecasting in Bangladesh
Created using ChatSlide
This study investigates machine learning temperature forecasting across 35 BMD stations from 1961 to 2023, addressing research gaps and objectives for nationwide predictions. It analyzes seasonal climatology and regional variability, achieving notable forecasting results with XGBoost and SARIMA models. Insights reveal that temperature history and seasonality are key factors, while additional weather predictors slightly improved accuracy. Statistical methods like Friedman and SHAP provide...