Yas-smine/Financial-Inclusion-in-Africa-Streamlit

# Financial Inclusion Prediction App in East Africa ## Objective This project aims to predict whether an individual in East Africa is likely to have or use a bank account, using demographic and socio-economic features. Financial inclusion means providing individuals and businesses with access to useful and affordable financial services in a responsible and sustainable way, such as payments, savings, credit, and insurance. ## Dataset The dataset contains information for approximately 33,600 individuals across East African countries (Kenya, Uganda, Tanzania, Rwanda) for the years 2016-2018. Key columns include: - `country`: Country of residence - `year`: Year of survey - `uniqueid`: Unique respondent ID - `bank_account`: Target variable (Yes/No) - `location_type`: Urban or Rural - `cellphone_access`: Yes/No - `household_size`: Number of people in household - `age_of_respondent`: Respondent's age - `gender_of_respondent` - `relationship_with_head` - `marital_status` - `education_level` - `job_type` ## Steps Followed 1. **Environment Setup** - Installed necessary Python packages: `pandas`, `numpy`, `scikit-learn`, `xgboost`, `imbalanced-learn`, `streamlit`, `joblib`. 2. **Data Exploration** - Loaded data with `pandas` and examined using `.info()`, `.head()`, `.describe()`. - Checked for missing values, duplicates, and outliers. - Explored value counts for categorical features. 3. **Data Preprocessing** - Simplified education and job categories to reduce sparsity. - Encoded categorical features using `LabelEncoder`. - Handled class imbalance using **SMOTE** (Synthetic Minority Oversampling Technique). - Scaled and transformed numeric features as needed. 4. **Model Training** - Trained **XGBoost classifier** on preprocessed data. - Evaluated with accuracy, precision, recall, and F1-score. - Achieved balanced performance after SMOTE. 5. **Streamlit App** - Created a web application with **Streamlit**: - Inputs for all demographic and socio-economic features. - Sim …