A type of Machine Learning where transactions are assumed to be valid initially and only later challenged if found fraudulent, aiming for faster processing on Layer 2 solutions.
