Constructive Potential Theory: Foundations and Applications.

Stochastic analysis is now an important common part of computing and mathematics. Its applications are impressive, ranging from stochastic concurrent and hybrid systems to finances and biomedicine. In this work we investigate the logical and algebraic foundations of stochastic analysis and possible applications to computing. We focus more concretely on functional analysis theoretic core of stochastic analysis called potential theory. Classical potential theory originates in Gauss and Poincare's work on partial differential equations. Modern potential theory now study stochastic processes with their adjacent theory, higher order differential operators and their combination like stochastic differential equations. In this work we consider only the axiomatic branches of modern potential theory, like Dirichlet forms and harmonic spaces. Due to the inherently constructive character of axiomatic potential theory, classical logic has no enough ability to offer a proper logical foundation. In this paper we propose the weak commutative linear logics as a logical framework for reasoning about the processes described by potential theory. The logical approach is complemented by an algebraic one. We construct an algebraic theory with models in stochastic analysis, and based on this, and a process algebra in the sense of computer science. Applications of these in area of hybrid systems, concurrency theory and biomedicine are investigated. Parts of this paper have been presented, in shorter form, at diverse conferences and workshops. This work represents a common 'umbrella' for all these presentations and offers an extended version for the (some time) very short published materials.