KathituCodes/Financial-Inclusion-in-Africa-ML-Classification-Project

# Financial Inclusion in Africa – ML Classification Project ````markdown ## Problem Statement Millions of individuals across East Africa remain excluded from formal financial systems. This project uses machine learning to predict whether a person is likely to own or use a **bank account**, based on their demographic and socio-economic attributes. ## Project Objective The primary goal is to: - Develop a machine learning classifier that predicts financial inclusion. - Identify underserved individuals and communities. - Enable smarter decision-making for financial outreach programs. ## Dataset The dataset contains information on **33,600 individuals** from East Africa, including: - Age, gender, education, and job type - Urban vs. rural location - Access to electricity, phone ownership, etc. **Target variable**: `bank_account` (Yes/No) ## Implementation Summary The project was implemented in the following phases: ### 1. Data Exploration & Cleaning - Displayed general information about the dataset - Used `pandas-profiling` for exploratory data analysis - Handled missing values, removed duplicates, treated outliers - Encoded categorical variables using label encoding ### 2. Model Training - Built and tested multiple classification models (e.g., Decision Tree, Random Forest) - Evaluated using cross-validation, accuracy, precision, and recall ### 3. Deployment via Streamlit - A local **Streamlit** app was built to interact with the model - The app accepts user input for all model features and provides real-time predictions - Trained model was saved using `joblib` and loaded into the Streamlit interface ## Results - The classifier successfully predicts bank account ownership with solid performance. - Streamlit interface allows real-time, user-friendly predictions. - The project supports data-driven interventions for improving financial inclusion. ## 🖥️ How to Use (Locally) 1. **Clone the repository**: ```bash git clone github.com …