Logic, Reasoning and A Programming Language for Simulating Economic and Business Processes with Artificially Intelligent Agents
The reason is that logics support the inference of true statements based on other inferences and axioms (which are also true statements). Thus these latter inferences are only implicitly encoded by the initial statements and the result only revealed to us explicitly (since we are, ourselves, beings of limited rationality) when the consequences of those statements in the logic are worked out. On the other hand, Bayesian processes state what adaptations shall be anticipated or, based on those anticipations, what actions shall be taken.
The true statements (axioms and inferences) naturally describe states so that the inferences from them are naturally taken to describe responses to those states. These responses are what we think of as processes. Bayesian models explicitly describe actions and the results of those actions (effectively processes) to yield what we think of as states. Thus, to a large extent a process modelled in a Bayesian way can not be a surprise to us, since it is specified more directly.
Generated with CERN WebMaker