In a MAN-MACHINE MODEL an actor or actors play roles while other parts of the model are implemented on a computer. A formal model has a
structure (the form of an equation, for example) and
parameters (the value of coefficients in an equation for example). Determination of both the structure and parameters is MODEL
identification ; determination of the parameters on the basis of experimental data is MODEL ESTIMATION. The check of a proposed model against experimental data other than those used for parameter estimation is model
validation . See also
verification . (IIASA)
A
system that stands for or represents another typically more comprehensive system. A model consists of a set of objects, described in terms of
variables and RELATIONs defined on these and either (a) embodies a theory of that portion of reality which it claims to represent or (b) corresponds to a portion of reality by virtue of an explicit
homomorphism or
isomorphism between the model's
parameters and given DATA. Four kinds of models are distinguished: (1) Sampling models consist of a mere subset of mutually exclusive objects from a larger universe of objects. The
representation is based on the assurance that each object of the universe had the same
probability to be included in the sample. (2) Iconic models (
see icon ) are
linear transforms of a configuration of objects in the universe. The representation is based on the assurance that an iconic model retains that universe's topological characteristics. E.g., scale models, photographs, the graphical representation of networks. In (3) behavioral models, the relations are
transformations , equations or operating
rules and the representation is based on the assurance that the
behavior of the model corresponds to the behavior of the
system modelled. This is established either by identifying the model's
parameters and equations, or showing that the homomorph is not violated. E.g., the computer
simulation of an economy, the model of a plane built into an automatic pilot. In a (4) symbolic model the set of objects are
symbols and the relations are expressed in the form of algebraic, computational or algorithmic statements exhibiting no behavior of its own. Symbolic models must be realized in or coupled with a machine in order to become a behavioral model of something else. E.g., a formal statement about a social process must be translated into the algorithmic form of a
program acceptable to a computer. Sampling models represent the materiality of the reality modelled. The other three do not. Their structural behavior or symbolic correspondence makes no reference to the physical nature of the objects represented in the model. (
Krippendorff )
