Fab2500/imanishimwefabrice_food-securityprediction

**food security levels across Rwanda** using **machine learning** and **geospatial analytics**. It provides **early-warning insights** to identify districts at risk of **food insecurity**, enabling proactive policy actions and disaster prevention. google drive full project link:drive.google.com # Rwanda Food Security Prediction & Early-Warning Dashboard ## Overview This project predicts and visualizes **food security levels across Rwanda** using **machine learning** and **geospatial analytics**. It provides **early-warning insights** to identify districts at risk of **food insecurity**, enabling proactive policy actions and disaster prevention. --- ## Dataset Used The system integrates data from: - **Comprehensive Food Security and Vulnerability Analysis (CFSVA) 2024** - **Seasonal Agriculture Survey 2024** These datasets contain key indicators such as food consumption scores (FCS), household demographics, crop productivity, and socio-economic variables. --- ## Background Food insecurity remains a significant challenge in Rwanda, especially in rural areas where communities depend on subsistence agriculture. Climate change, crop failure, and socio-economic vulnerabilities can quickly trigger hunger crises. Traditional surveys like the **CFSVA (2024)** and **Seasonal Agriculture Survey (2024)** provide valuable insights but are **retrospective**. This project builds a **data-driven predictive system** to: - Enable **proactive monitoring** - **Reduce food scarcity risks** - **Support national and NGO-level decision-making** --- ## Project Goal To develop an **AI-powered Food Security Early-Warning Dashboard** that predicts the likelihood of **food insecurity in Rwandan districts** using agricultural, environmental, and socio-economic data. ### Specific Objectives 1. **Data Integration** – Merge CFSVA and Agriculture Survey datasets. 2. **Feature Analysis** – Identify critical determinants of food insecurity (e.g., crop yield, household size, income). 3. **Predictive Modeling** – Build a **Random Forest ML model** to estimate district-level risk. 4. **Visualization** – Design an **interactive Streamlit dashboard** with maps, charts …