The digital world is growing rapidly. We are used to performing many of our daily tasks online, such as booking cabs, shopping on e-commerce websites, and even recharging our phones. For the majority of these tasks, we are used to paying with credit cards. However, it is a known fact that a credit card can be compromised, which could result in a fraudulent transaction. The Nilson report estimates that for every $100 spent, seven cents are stolen. It estimates the total credit card fraud market to be around $30 billion.
Detecting whether a transaction is fraudulent or not is a very impactful data science problem. Every bank that issues credit cards invests in technology to detect fraud and take the appropriate actions immediately. There are lot of standard supervised learning techniques such as logistic regression, from random forest...
