Since technology has expanded and evolved, fraud in the financial industry has developed considerably, costing firms and consumers hundreds of billions of dollars each year. The fraudsters who target the financial sector are in a fast phase of adapting to techniques to exploit gaps in current protective measures. These crimes include insider trading, money laundering, credit card fraud, stock and commodity fraud, and health and auto insurance fraud. Systems for combating fraud alone do not offer enough protection against these crimes. As a consequence, it is more critical than ever to have fraud detection systems in place to detect fraudulent acts after they have occurred and perhaps save money. Over the past several decades, researchers have intensively investigated anomaly detection approaches for this purpose; many of these methods have used statistical, artificial intelligence, and machine learning models. Until recently, the most prevalent models studied in research were supervised learning algorithms.
