Cognitive neural network for cybersecurity

This chapter discusses the future of cybersecurity as warfare between machine learning techniques of attackers and defenders. As attackers will learn to evolve new camouflaging methods for evading better and better defenses, defense techniques will in turn learn new attacker´s tricks to defend against. The better technology will win. Here we discuss theory of machine learning based on dynamic logic that are mathematically provable to learn with the fastest possible speed. We also discuss cognitive functions of dynamic logic and related experimental proofs. This new mathematical theory, in addition to being provably fastest machine learning technique, is also an adequate model for several fundamental mechanisms of the mind.