https://github.com/jbeno/credit

Analysis of German credit risk data
https://github.com/jbeno/credit

Last synced: 9 days ago
JSON representation

Analysis of German credit risk data

• Host: GitHub
• URL: https://github.com/jbeno/credit
• Owner: jbeno
• Created: 2023-07-11T17:31:33.000Z (over 1 year ago)
• Default Branch: main
• Last Pushed: 2023-07-11T18:13:35.000Z (over 1 year ago)
• Last Synced: 2024-10-06T15:41:47.176Z (about 1 month ago)
• Language: Jupyter Notebook
• Size: 754 KB
• Stars: 1
• Watchers: 1
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

        
# credit

This is an analysis of the [Statlog (German Credit Data)](#https://archive.ics.uci.edu/dataset/144/statlog+german+credit+data) from the UCI Machine Learning Repository. It was done as a project for the U.C. Berkeley Certificate in Machine Learning & Artificial Intelligence. The entire process is documented in this Jupyter notebook file:

* [credit.ipynb](credit.ipynb)

## Summary of Findings

Here are the key insights from my analysis:

* The majority of customers have a **"good" credit class** (70%)

* Almost all are **foreign workers!** (96.3%) I do believe this dataset came from Germany

* Most are **male** (69%) **skilled workers** (63%) that **own a home** (71.3%) and have just **1 dependent** (84.5%), themselves

* The **women are younger than the men**, with most between the ages of 20 and 30

* The most popular purpose is for a **Radio or TV** (28%), followed closely by a **new car** (23.4%)

* However, **Females** obtained a greater proportion of loans for **furniture/equipment, education, and domestic appliance** vs. males

* Most customers (60.3%) have a **Savings** with **less than 100 Deutsche Mark** (1 DEM = $0.55 USD)! No wonder they are applying for credit

* The **average credit amount** was 3,271 DEM ($1,825 USD)

* The **largest credit amount** was for the **Other** category or purpose (6,948 DEM median), followed by **Used Car** (4,788 DEM median)

* **Credit Amount vs. Duration** have the most positive correlation, which is strong at 0.62

* **Credit Amount vs. Installment Commitment** have the most negative correlation, which is weak at -0.27

## Helper Functions

As part of this project, I created some helper functions that are published here:

* [mytools.py](https://github.com/jbeno/mytools)