Description vs. Modelling
暂无分享,去创建一个
Summary In the area of Information and Systems recent research devoted to complex systems has emphasized mathematical modelling usually implemented on digital computers by procedural control structures. For some classes of systems insufficiently known or where input contains unpredictable elements this approach can lead to quite complicated models hardly tractable as a set of conventional programs. On the other hand research in Artificial Intelligence has produced new data structures and non procedural techniques for dealing with systems whose laws are imperfectly known and fragmentary, the emphasis being set upon reasonably efficient implementations. This paper will advocate the use of the latter techniques, designated as the descriptive approach . The major points to be discussed are: what is a descriptive approach and how does it differ from modelling? How is it related to simulation? What type of problems can it solve? How is it implemented? What are the forseeable developments? Although all those questions do not have a definite and final answer, data description is an important issue when speaking of Information and Systems. In particular it allows one to implement very easily asynchronous, non hierarchical, interacting processes, whose decisions are reached by a consensus resulting from an equilibrium between the various independent parts.
[1] G. Powers,et al. Synthesis of system designs: III. Toward a process concept generator , 1971 .
[2] D. H. Smith,et al. Applications of artificial intelligence for chemical inference. 8. An approach to the computer interpretation of the high resolution mass spectra of complex molecules. Structure elucidation of estrogenic steroids. , 1972, Journal of the American Chemical Society.
[3] Randall Davis,et al. An overview of production systems , 1975 .