Esthern-Muigai/AI-Yield-Predictor-In-Kenya

# AI-Powered Crop Yield Prediction for Enhanced Food Security in Kenya ## Project Overview This project leverages Artificial Intelligence and Machine Learning to predict crop yields (specifically Maize, a staple crop) for smallholder farmers in Kenya. By providing accurate, hyper-local yield forecasts, we aim to enhance food security, improve farmer livelihoods, and enable data-driven agricultural planning across the nation. This initiative directly addresses **UN Sustainable Development Goal (SDG) 2: Zero Hunger**, by empowering farmers with timely insights to optimize planting, manage resources efficiently, and make informed market decisions. ## Problem Statement Smallholder farmers in Kenya face significant challenges due to unpredictable climate patterns, limited access to accurate agricultural data, and susceptibility to pests and diseases. This uncertainty leads to sub-optimal resource allocation, reduced yields, and increased vulnerability to food insecurity. This project seeks to mitigate these challenges by providing proactive yield predictions. ## Machine Learning Approach Our solution employs a multi-modal, hybrid AI approach: * **Supervised Regression (RandomForestRegressor):** The core model for predicting crop yield, trained on historical and environmental data. * **Time Series Forecasting (Conceptual: LSTMs/Transformers):** In a real-world application, a time-series model would forecast future weather patterns (rainfall, temperature) to provide proactive inputs. * **Computer Vision (Conceptual: CNNs):** In a real-world application, a CNN would analyze satellite imagery for crop health (NDVI) and stress detection, providing additional features. * **Ensemble Learning:** The RandomForestRegressor itself is an ensemble method, combining multiple decision trees for robust predictions. ## Data Sources (Simulated for Prototype) For this prototype, we've simulated data to demonstrate the concept, reflecting typical Kenyan agricultural contexts. In a real …