Non-conventional computing paradigms

Abstract

The conventional computing paradigms are usually two: the symbolic representational one and the connectionist (or mechanistic) approach. The first one is dominant since the Christening of Artificial Intelligence in the Summer of 1956. This paradigm assumes that everything that has to be computed needs previously to be represented in terms of data structures and inferential schemes, including algorithms and implementation programs. So, all the information needs to be explicit from the beginning in a declarative manner. The second paradigm, usually known as connectionist, includes the field of Artificial Neural Networks, and other biologically inspired approaches. In essence the first paradigm is related to General Purpose von Neumann Architecture, while the second one is related to Special Purpose Machines in which the control is explicit in the specific electronic circuit that embodies each information processing machine. In addition to these two dominants paradigms, there are other new approaches to the task of computation, not so well established, but in the brainstorming frontier of the science and engineering. These new approaches include Quantum Computation, Membrane Computation, Evolutionary Algorithms and many other Conceptual Proposals still far away from feasible and competitive implementations. This special issue of Natural Computing deals with some of these new paradigms. The articles are extended versions of papers selected from the second International Work Conference on the Interplay between Natural and Artificial Computation (IWINAC) held in La Manga del Mar Menor (Spain) during June 2007.