原始名称记载: UNIVERSITY OF SOUTH FLORIDA

Systems and methods are provided herein for advancing ensemble learning methods, including stacking, boosting, and bagging, for defeating data protection approaches by converting their generated unlearnable data into learnable ones. Processes of the present disclosure may enhance and implement ensemble learning on the unlearnable data while incorporating nonlinear transformations.

A method for generating a digital signature for a message involves obtaining the message and a signing private key. The signing private key is divided into a plurality of indexed values. A counter value is identified to provide a set of partitions with unique integer interpretations through at least one round of combining the message with the counter value, hashing the combined message and counter value to generate a hash, reducing the hash to a predetermined length to generate a reduced hash, and partitioning the reduced hash into a plurality of partitions. If the partitions are not interpretable as a set of unique integers, the counter value is incremented for the next round. If the partitions are interpretable as a set of unique integers, the counter value is provided. A set of values from the signing key indexed by the unique integers is selected, and the digital signature comprising the selected set of values and the counter is outputted.