ML Bank Marketing Project

Machine Learning Project to build a prediction model to score customer propensity to subscribe bank products.

Project Data: https://archive.ics.uci.edu/ml/datasets/Bank+Marketing#

Methodology: The Customer Propensity ML use case was developed using the CRISP-DM methodology.

The repository is divided into different directory:

• bin: The bin directory contains the ML Pipeline written in Python. The pipeline automates the feature engineering and model development using sklearn pipeline package. We can re-use the source code to convert it into a AWS Sagemaker Model or deploy the pipeline for real time applications such as AWS Lambda
• dev_notebook: The directory contains the Dev Jupyter notebook featuring Exploratory Data Analysis, Model Selection and building. The univariate analysis html file was generated by pandas-profile package to analize data, correlation and interaction. The TOC contents of the dev notebook highlights the steps taken while building the ML Solution.
• datasets - Contains the Raw data and the train/test split data