Credit Card Customer Dataset Analysis

Dataset Purpose

The experiment aims to perform preliminary data analysis on the credit card customer dataset in the security domain, leveraging Python frameworks and libraries.

Analysis Objectives

1. Explore the credit card customer dataset and compute main statistical indicators.
2. Construct various dependency relationships among existing attributes in the dataset.
3. Visualize analysis results using multiple chart types.

Analysis Questions

We will attempt to answer a set of questions that may be relevant when analyzing credit card customer data:

1. What is the average age of all customers?
2. How many customers have a payment default issue next month across the whole dataset?
3. How many married and single customers are there?
4. What percentage of variables are non‑unique?
5. Based on the complete correlation matrix, which values exhibit very high dependency?
6. How can we identify new trends in the customer operation roadmap?
7. How do we define the boundary where most data points of feature pairs are densely clustered?
8. What proportion of customers are likely to default on payment next month in our DataFrame?
9. Among the attracted customers, what are the average values of numerical features (average age and average repayment status delay)?
10. For typical customers likely to default next month, what was the average repayment status in September 2005?
11. How many customers paid on time in September 2005?

Libraries Used

import pandas as pd
import matplotlib.pyplot as plt
import numpy as np
import seaborn as sns

Dataset Download

!wget https://archive.ics.uci.edu/ml/machine-learning-databases/00350/default%20of%20credit%20card%20clients.xls
!mv -f default of credit card clients.xls CreditCard.xls

Result Visualization Examples