Learning-Based Cryptography

Abstract

Machine learning technologies have seen a broad variety of applications in recent years. Nonetheless, the use of machine learning in the fields of information and network security is not a discovery. Machine learning and cryptography have some commonalities. One of the most noticeable elements is the processing of large amounts of data and large search areas. Machine learning, with its broad set of approaches, has emerged as an enthralling field of study with enormous potential for practical application. Machine learning approaches, both supervised and unsupervised, have been used in various data-security-related applications during the past thirty years, including cryptographic algorithms, cryptanalysis, and steganography.