Towards soft computing

Abstract

It is not clear who coined the term soft computing but one of its greatest promoters has been none other than Lotfi Zadeh — the inventor of fuzzy sets theory. He sees soft computing as an extension of fuzzy logic by merging it with neural networks, (or neurocomputing — NC as he puts it) and evolutionary computing. He should certainly be acknowledged as the earliest founders of soft computing which he promoted as head of the Berkeley Initiative in Soft Computing — the BISC group. Zadeh advocates soft fomputing as the means by which we may go beyond what AI has been able to achieve during its 40 years' existence. In this paper we try and look more deeply into what Zadeh has to say about both the motivation for and the methodology of soft computing.

This paper was given as a keynote address at the Expert systems 95 conference and first appeared in the proceedings of that conference. It is reprinted here with their permission.