How mohtadaayesha/bank-marketing works
This repository is a notebook-based bank-marketing classification project. Its principal workflow takes the documented Bank Marketing dataset through inspection, exploration, target preparation and stratified splitting; preprocesses features and trains a class-balanced Logistic Regression model; then evaluates and interprets the model and demonstrates a new-customer prediction. The README documents these stages, but no notebook source excerpts were available, so the graph reflects documented relationships rather than verified implementation details.
Main entry points: Bank_Marketing.ipynb Feature preprocessing
Components 16
Data exploration
- Inspect and quality-check
Bank_Marketing.ipynb - Explore features and target
Bank_Marketing.ipynb
Data preparation
- Encode subscription target
Bank_Marketing.ipynb - Stratified data splits
Bank_Marketing.ipynb - Feature preprocessing
Bank_Marketing.ipynb - Impute missing values
Bank_Marketing.ipynb - Scale numerical features
Bank_Marketing.ipynb - Encode categorical features
Bank_Marketing.ipynb
Modeling
- Train balanced Logistic Regression
Bank_Marketing.ipynb - Predict subscription
Bank_Marketing.ipynb
Results and prediction
- Evaluate test predictions
Bank_Marketing.ipynb - Interpret coefficients
Bank_Marketing.ipynb - New-customer prediction
Bank_Marketing.ipynb
Other components
- Bank Marketing data
- New customer
- Yes / no result
Connections 15
- Bank Marketing data → Inspect and quality-check: loads for checks
- Bank Marketing data → Explore features and target: explored in notebook
- Bank Marketing data → Encode subscription target: provides target
- Encode subscription target → Stratified data splits: precedes splitting
- Stratified data splits → Feature preprocessing: provides feature sets
- Feature preprocessing → Impute missing values: applies transformations
- Feature preprocessing → Scale numerical features: applies transformations
- Feature preprocessing → Encode categorical features: applies transformations
- Feature preprocessing → Train balanced Logistic Regression: pipelines with
- Train balanced Logistic Regression → Predict subscription: fits model for
- Predict subscription → Evaluate test predictions: supplies test outputs
- Train balanced Logistic Regression → Interpret coefficients: provides coefficients
- New customer → New-customer prediction: provides features
- New-customer prediction → Predict subscription: uses trained pipeline
- New-customer prediction → Yes / no result: returns prediction