Fraud detection on remote banking: Unusual behavior on historical pattern and customer profiling

The goal of this thesis is mentioning about history of informatics crimes and fraud on online banking systems. Additionally, I am going to talk about fraud detection systems, the types of these systems and detection models. Banks use specialized software solutions in order to avoid fraud issues as well as other sector companies. Banks's loses are unpredictably higher than the others which means they need to be consistent, determined, fast and smarter than fraudsters. The thesis is going to talk about the architecture of these systems, how data is evaluated internally and how the output is important for banks. Moreover, I will talk about the customer profiling technique that is a machine learning technique which is more precise and makes output easier to handle. Without that technique, all software solutions would be just collection of rules and became just a 'machine'. However, we need a machine that is smarter than a human. Criminals do not follow rules. Why should we?