Mrisaac7979/crop-recommendation-togo

# Crop Recommendation System - Togo Agro-climatic Zones Multi-class machine learning pipeline for smallholder crop advisory in Togo. Given soil properties, climate variables, and agro-climatic zone, the model recommends the most suitable crop among five staple crops cultivated in the country. --- ## Problem Statement Smallholder farmers in Togo face yield losses partly due to suboptimal crop selection relative to local soil and climate conditions. This project frames crop recommendation as a supervised multi-class classification problem, using agro-climatic features as predictors. This work is developed as a methodological prototype in preparation for graduate research on precision agriculture and climate-adaptive farming systems in West Africa. --- ## Crops and Agro-climatic Zones **Target classes:** Maize, Cassava, Sorghum, Yam, Soybean **Regions modeled:** | Zone | Mean Rainfall (mm/yr) | Mean Temp. (C) | Dominant crops | |-----------|----------------------|----------------|-----------------------| | Savane | 750 | 29.5 | Maize, Sorghum | | Kara | 1050 | 27.5 | Maize, Cassava | | Centrale | 1100 | 27.0 | Maize, Yam, Cassava | | Plateaux | 1300 | 25.5 | Cassava, Yam | | Maritime | 1450 | 27.0 | Cassava | --- ## Features | Feature | Unit | Description | |----------------|-----------|------------------------------------------| | rainfall_mm | mm/year | Annual cumulative rainfall | | temp_mean_c | C | Mean annual temperature | | soil_ph | - | Soil pH (4.5 - 8.0) | | soil_n_ppm | ppm | Soil nitrogen content | | soil_p_ppm | ppm | Soil phosphorus content | | soil_k_ppm | ppm …