ABSTRACT
Cryptographic systems based on elliptic curves have been introduced as an alternative to conventional public key cryptosystems. The security of both kinds of cryptosystems relies on the hypothesis that the underlying mathematical problems are computationally intractable, in the sense that they cannot be solved in polynomial time. Artificial neural networks are computational tools, motivated by biological systems, which have the inherent ability of storing and making available experiential knowledge. These characteristics give to the artificial neural networks the ability to solve complex problems. In this paper, we study the performance of artificial neural networks on the approximation of data derived from the use of elliptic curves in cryptographic applications.
