larrymabote-sk/africa-gdp-growth-ml

Machine Learning project predicting GDP growth across selected African economies using World Bank data. # Africa GDP Growth Prediction ## Project Overview This project applies Machine Learning to investigate whether selected economic indicators can be used to predict annual GDP growth across selected African economies. The project follows an end-to-end Machine Learning workflow, including data collection, preprocessing, exploratory data analysis, model development, evaluation, and deployment. ## Problem Statement GDP growth varies considerably across African economies and is influenced by multiple economic factors. This project investigates whether indicators related to infrastructure, investment, industrial activity, and financial development contain useful information for predicting GDP growth. The project is intended as an exploratory predictive analysis rather than a causal economic model. ## Dataset The data was obtained from the World Bank Open Data API. The analysis covers Ghana, Kenya, Mozambique, Nigeria, and South Africa from 2005 to 2024. The target variable is: * GDP Growth (%) The main predictor variables are: * Electricity Access (% of population) * FDI Inflows (% of GDP) * Industry Share of GDP (%) * Gross Capital Formation (% of GDP) * Domestic Credit (% of GDP) For the extended modelling experiment, Nigeria was excluded because the additional indicators were unavailable for the required period. One incomplete Kenya observation was also removed, resulting in 79 complete observations. ## Methodology The project followed these main stages: 1. Data collection from the World Bank API 2. Data cleaning and preprocessing 3. Missing-value and duplicate checks 4. Exploratory Data Analysis 5. Correlation and trend analysis 6. Train-test split 7. Linear Regression modelling 8. Random Forest Regression modelling 9. Model evaluation using MAE, RMSE, and R² 10. Deployment using Streamlit ## Model Results Linear Regression performed better than Random Forest on the initial test set. The extended Linear Regression model, using all five economic ind …