Case Study: Credit Card Transactions Analysis

[Pankaj-Str]

📌 Case Study: Credit Card Transactions Analysis

Objective:

This case study will help you develop expertise in data cleaning, manipulation, analysis, and visualization using Pandas. You will analyze a dataset of credit card transactions to detect trends, find anomalies, and extract insights that could help in fraud detection and customer behavior analysis.

---

📂 Problem Statement:

You work as a data analyst for a financial company that provides credit cards to customers. Your task is to analyze credit card transactions and provide insights that can help in decision-making.

You have been provided with a dataset named credit_card_transactions.csv, which contains details about transactions made by various customers.

---

📊 Dataset Overview (credit_card_transactions.csv)

Download Dataset

Column Name	Data Type	Description
Transaction_ID	int	Unique transaction identifier
Customer_ID	int	Unique customer identifier
Transaction_Date	str	Date of transaction (YYYY-MM-DD)
Transaction_Type	str	Type of transaction (Online, POS, ATM, etc.)
Merchant	str	Name of the merchant
Category	str	Purchase category (Groceries, Electronics, Travel, etc.)
Amount	float	Transaction amount (in USD)
Payment_Mode	str	Payment method (Credit Card, Debit Card, etc.)
Transaction_Status	str	Status (Approved, Declined, Pending)
Location	str	City where transaction occurred

---

📝 Assignment Questions:

1️⃣ Data Exploration and Cleaning

• Load the dataset into a Pandas DataFrame and display the first 5 rows.
• Check the shape, column names, and summary statistics of the dataset.
• Identify and handle missing values (fill or drop based on the data type).
• Convert Transaction_Date into datetime format and extract year, month, and day as new columns.

---

2️⃣ Data Selection and Indexing

• Retrieve all transactions made in January 2024.
• Find transactions where Amount > 1000 and Transaction_Type is "Online".
• Select only Approved transactions from the dataset.

---

3️⃣ Data Manipulation and Feature Engineering

• Create a new column Discounted_Amount, assuming a 5% discount on all transactions above $500.
• Categorize the Transaction_Amount into Low, Medium, and High based on:
  • Low: Below $100
  • Medium: Between $100 - $500
  • High: Above $500
• Drop the Merchant column if more than 30% of values are missing.

---

4️⃣ Aggregation and Insights

• Find the total transaction amount per Category.
• Determine the number of declined transactions per Payment_Mode.
• Identify the top 5 most frequent merchants based on transaction count.
• Find the average transaction amount per Location.

---

5️⃣ Fraud Detection Indicators

• Find customers who made more than 10 transactions in a single day (potential fraud).
• Identify transactions that have the same Customer_ID but occurred in different locations within 5 minutes.
• Find transactions where Amount > $5000 and Transaction_Type is Online (flag as high-risk).

---

6️⃣ Data Merging and Joining

• Suppose you have another dataset (customer_info.csv) containing Customer_ID, Age, Gender, and Account_Status.
  • Merge it with credit_card_transactions.csv using an appropriate join operation.
  • Find the average transaction amount per Age group.

---

📊 Bonus Challenge (Optional)

• Create a bar chart showing the total transaction amount per Category using Matplotlib or Seaborn.
• Generate a heatmap showing correlation between Amount, Transaction_Status, and Payment_Mode.

---

📌 Submission Guidelines:

1. Write your Pandas code with comments explaining each step.
2. Provide brief answers to descriptive questions.
3. Submit your Jupyter Notebook (.ipynb) or Python script (.py) along with a short summary report (if needed).

---

✅ Evaluation Criteria:

✔ Correctness: Ensure your solutions are accurate and functional.
✔ Code Readability: Use proper variable names and comments.
✔ Data Handling: Efficiently handle missing values and incorrect data types.
✔ Fraud Detection & Insights: Extract meaningful business insights.

🚀 Good luck! Happy Coding!

---

[ritvik-patil]

https://github.com/ritvik-patil/Credit-Card-Transactions-Analysis

[GawliYash20]

https://github.com/GawliYash20/Case-Studies-Data-Science/blob/main/Credit_card_transaction.ipynb