Meta nite Model Theory
نویسندگان
چکیده
Motivated by computer science challenges, we suggest to extend the approach and methods of nite model theory beyond nite structures. We study de nability issues and their relation to complexity on meta nite structures which typically consist of (i) a primary part, which is a nite structure, (ii) a secondary part, which is a (usually in nite) structure that can be viewed as a structured domain of numerical objects, and (iii) a set of \weight" functions from the rst part into the second. We discuss model-theoretic properties of meta nite structures, present results on descriptive complexity, and sketch some potential applications.
منابع مشابه
ON THE USE OF KULSHAMMER TYPE INVARIANTS IN REPRESENTATION THEORY
Since 2005 a new powerful invariant of an algebra has emerged using the earlier work of Horvath, Hethelyi, Kulshammer and Murray. The authors studied Morita invariance of a sequence of ideals of the center of a nite dimensional algebra over a eld of nite characteristic. It was shown that the sequence of ideals is actually a derived invariant, and most recently a slightly modied version o...
متن کاملTHEORIES IN FINITE MODELTHEORYJohn
Finite model theory stands at the intersection of complexity theory and model theory. Most of the work thus far in nite model theory is analogous to work in 1950's style model theory of rst order logic. That is, the focus is on the study of properties of the logic: expressibility, preservation theorems, criteria for equivalence among models. Around 1960, with Vaught's paper 22], which is still ...
متن کاملObservant: An Annotated Term-Rewriting System for Deciding Observation Congruence
The use of term-rewriting systems (TRSs) is at the core of automated theorem proving (ATP). There are a number of methods whereby we can build terminating TRSs with full deduction power, with respect to some theory. Yet, there exist theories these techniques cannot handle, giving rise to a trade-oo between deduction power and termination. One such theory is the theory of observation congruence ...
متن کاملMetalevel Learning -is More than Model Selection Necessary?
This workshop paper proposes an alternative to the techniques commonly regarded as met-alevel learning. Recapitulating results from statistical learning theory regarding selection of learning algorithms, we cannot guarantee with probability one to select the best of n methods on a nite sized validation set. Hence selecting a learner is diicult. What makes meta level learning even more diicult i...
متن کاملDatabases and Finite-Model Theory
Databases provide one of the main concrete scenarios for nite-model theory within computer science. This paper presents an informal overview of database theory aimed at nite-model theorists, emphasizing the speciicity of the database area. It is argued that the area of databases is a rich source of questions and vitality for nite-model theory.
متن کامل