The main results expected from our studies will be published in reports devoted to questions like the following: The investigation of the relationship between discrete and continuous models of computation (processes). Better understanding of the topology of a space of algorithms, e.g. for answering questions on stability. Development of similarity notions (measures) for algorithms and proofs (essential for certain aspects of re-use and approximative analysis). Optimization of programs or proofs (like with neural networks) instead of construction or search (like in logic programming). A mathematical model for reasoning about probabilistic algorithms (leaving of the common dichtonomy: determinism vs. nondeterminism).
More practical outcomes shall be gained by applying the theoretical results obtained for introducing measures for the quality and optimality of software (not just referring to correctness but also to stability, re-usability, error probability, etc.) aiming towards an approximative analysis of programs, and the development of methods and tools for the initialization and knowledge extraction for general neural networks.