Lilianigwegbe/AI-DEVS-SDG1-Africa-Poverty-Forecast-ml

# 🌍 Africa Poverty Forecasting App ## 📌 SDG 1: No Poverty --- ## 📖 About This Project This is AI software engineering project was created as part of our Sustainable Development Goals (SDG) machine learning assignment. We chose **SDG 1: No Poverty**, focusing on forecasting poverty trends in African countries. By using a machine learning model, we aim to predict how poverty rates might change over time — helping governments, communities, and organizations plan better for the future. 🌱 --- ## 🎯 Problem We Are Solving Poverty remains one of Africa’s biggest challenges. Knowing how poverty levels might change in the coming years can help decision-makers take action sooner. Our app makes this easier by: - Forecasting poverty rates for African countries - Showing both past and future trends in simple, clear charts - Helping users explore data in a clean, interactive way --- ## 🧠 Machine Learning Approach ### 📌 What Technique Did We Use? We used a **Supervised Learning** method called **XGBoost Regression**. **What is XGBoost?** It’s a popular, beginner-friendly machine learning algorithm known for being fast and accurate. It works by combining simple prediction models (called decision trees) into a more powerful one. **How it works in our project:** - We trained the model using poverty data from past years - Created lagged features (like previous years’ poverty rates) to help the model learn trends over time - Used the model to predict future poverty rates for up to 15 years ahead --- ## 📊 Dataset We used a cleaned dataset called **`africa_poverty_cleaned.csv`**, which includes: - Country names - Years - Poverty rates (as a percentage) The data was sourced from publicly available SDG databases and cleaned for analysis. --- ## 🛠️ Tools and Libraries This project was built using: - **Python 3.10** - **Streamlit** (for creating the interactive web app) - **Pandas** (for data manipulation) - **XGBoost** (for regression modeling) - **Matplotlib** (for ch …