Fraud & Risk
Machine Learning: The Secret Ingredient for Detecting and Preventing Fraud
With online fraud showing no signs of abating anytime soon, many online businesses are finding that rules-based approaches are not scaling to handle increasingly sophisticated fraud patterns and datasets, and are turning to machine learning to detect suspect behavior and prevent illegitimate transactions on their platforms. This whitepaper provides a comparison between machine learning and rules-based detection systems, then explains how machine learning actually works. The paper then compares and contrasts supervised and unsupervised machine learning models, and includes real-world examples to help illustrate the benefits and drawbacks inherent with each approach.