Cybercriminals have built all kinds of new tools and techniques specifically intended to evade these adaptive solutions. Traditional risk engines use a pure statistical model for risk scoring, such as Nave Bayesian, Decision Trees and Neural Networks. These statistical models rely heavily on device identification as their detection method. As threats have become more sophisticated and bypassing two-factor authentication has become common for cybercriminals, stopping fraud now requires more decisive action. An evidence-based detection solution is founded upon that is founded on visibility of criminal attacks.”]
Source: https://securityintelligence.com/fraud-risk-engines-the-statistics-are-not-in-your-favor/