karuribrian/agriculture-ml-predictive-analytics

Machine learning models for predicting crop yield and rainfall trends in Kitui County, Kenya. Includes Random Forest, XGBoost, and ARIMA models with insights for climate-resilient agriculture. # 🌾 Climate-Smart Agriculture in Kitui County **Data-driven crop yield prediction and rainfall forecasting using machine learning.** ## 📌 Project Overview This project applies predictive analytics to agriculture in Kitui County, Kenya — a region facing climate challenges like erratic rainfall and soil degradation. We use machine learning models to: - Predict annual crop yields based on environmental factors - Forecast rainfall trends to anticipate drought risk - Recommend smart, climate-resilient farming practices ## 🧠 Objectives 1. **Predict Crop Yield:** Using Random Forest and XGBoost to forecast yields based on rainfall, soil moisture, temperature, and fertilizer use. 2. **Forecast Rainfall:** Using ARIMA to model and project rainfall patterns up to 2035. 3. **Actionable Insights:** Recommending drought-tolerant crops, smart irrigation, and data-informed decisions. ## 📊 Tools & Technologies - Python - Jupyter Notebook - pandas, scikit-learn, xgboost, matplotlib, seaborn, statsmodels, pmdarima - Machine Learning + Time-Series Forecasting ## 📁 Files Included - `Kitui_Crop_Yield_Simulation.csv`: Sample dataset with environmental & agricultural data - `kitui_crop_yield_forecast.ipynb`: Jupyter notebook with code, models, and visualizations - `README.md`: This documentation ## 🚜 Insights - Top predictors of yield: rainfall, soil moisture, crop type - Forecast suggests a 10–15% decline in rainfall by 2035 - Recommends drought-resilient crops like pigeon peas and sorghum, plus smart irrigation ## 💬 How to Use Clone the repo and run the Jupyter Notebook: ```bash git clone github.com cd kitui-crop-forecast jupyter notebook