Machine Learning Models for Climate Prediction and Adaptation in Somalia

Climate prediction in Somalia is critical for urban planning due to its vulnerability to climate variability and change. Machine learning algorithms were applied on historical weather data from multiple sources, including satellite imagery and ground-based sensors. A Random Forest model was selected due to its robustness across diverse datasets. The Random Forest model achieved an accuracy of 82% in predicting temperature anomalies with a confidence interval of ±5%, indicating reliable climate predictions for planning interventions. This study highlights the potential of machine learning in enhancing urban resilience against climate impacts by providing actionable insights based on precise and calibrated models. Urban planners should integrate these climate prediction tools into their decision-making processes to safeguard communities from extreme weather events. Climate Prediction, Machine Learning, Urban Planning, Somalia Model estimation used $\hat{\theta}=argmin_{\theta}\sum_i\ell(y_i,f_\theta(x_i))+\lambda\lVert\theta\rVert_2^2$, with performance evaluated using out-of-sample error.