Overview
Credit Card Fraud is one of the frauds which can happen both offline and online.From someone stealing your credit card to credit card number, hackers are always ready to breach in this kind of information.
That is why credit card companies suffer from churn as people do not feel safe after one incident.
Business challenge
Fraud detection is a very challenging problem.
The main reason is that fraudulent transactions are rare, they represent a very small fraction of total transactions within an organization.
The Credit Card Fraud Detection Problem includes analyzing past credit card transactions with the ones that turned out to be fraud.
Solution Offered
Fraud Detection model is based on Random Forest algorithm which takes hundreds of decision trees and aggregates and to build the model that tends to avoid overfitting due to class imbalance.
For this, Analysis of at least 20-50 last transactions of the customer is done to identify an underlying pattern.
Python and Machine Learning were used to build the model.