JO-ANN-BIRUNGI/DATA-SCIENCE-AND-MACHINE-LEARN-CAPSTONE-PROJECT

Forecasting Daily Air Temperature in Doho, Uganda Using Machine Learning and Deep Learning Approaches # DATA-SCIENCE-AND-MACHINE-LEARN-CAPSTONE-PROJECT Forecasting Daily Air Temperature in Doho, Uganda Using Machine Learning and Deep Learning Approaches This project aimed to forecast daily air temperature in Doho, Uganda, using historical weather data from multiple stations (2013–2016). Accurate temperature predictions are essential for agriculture, energy planning, and disaster preparedness, where reliable forecasts inform critical operational decisions. The dataset included daily measurements of air temperature, humidity, minimum and maximum temperatures, and station identifiers. Feature engineering was applied to capture temporal and seasonal patterns, including lag variables (1–14 days), rolling statistics (7-day averages and standard deviations), date-based features (month, day-of-year, weekday), and one-hot encoded station identifiers. Two modeling approaches were employed: Random Forest Regressor (RF): Classical machine learning model capturing nonlinear relationships. Long Short-Term Memory (LSTM): Sequential deep learning model capable of learning temporal dependencies. Model performance was evaluated using Mean Squared Error (MSE), Mean Absolute # Recommendations for Stakeholders Farmers and Agricultural Planners Use the daily temperature forecasts to optimize irrigation schedules and crop protection strategies. Align planting and harvesting decisions with predicted temperature trends to reduce crop stress and increase yield. Combine temperature forecasts with rainfall and humidity data for precision agriculture. Energy Providers Incorporate forecasts into demand planning, as electricity consumption often varies with temperature. Use predictions to manage load distribution and prevent outages during temperature extremes. Plan maintenance of energy infrastructure during predicted mild temperature periods. Disaster Management Authorities Utilize forecasts to anticipate heatwaves or cold spells, improving early warning systems. Prepare communitie …