Selected aspects of the calculus of self-modifiable algorithms theory

Abstract

In this paper the introductory concepts and chosen applications of the Calculus of Self-modifiable Algorithms (CSA) are presented. The CSA model is a theory for describing parallel behavior. In contrast to the well-known Hoare’s CSP and Milner’s CCS parallel computing theories, the CSA model was designed as a theory within Artificial Intelligence. It considers self-modifiable algorithms which are mathematical models of processes with the possibility of applying modifications to their own behavior. The CSA model unifies artificial intelligence methods and parallelism, two fundamental aspects of new generation computers. The most important features of the CSA model are the self-modifiability of programs (changeability of code) and the optimization of control (minimizing cost of algorithm execution).