We present an overview of four approaches of the finite automata use in stringology: deterministic finite automaton, deterministic simulation of nondeterministic finite automaton, finite automaton as a model of computation, and compositions of finite automata solutions. We also show how the finite automata can process strings build over more complex alphabet than just single symbols (degenerate symbols, strings, variables).
This article is the second in a series of two Mizar articles constituting a formal proof of the Gödel Completeness theorem [15] for uncountably large languages. We follow the proof given in [16]. The present article contains the techniques required to expand a theory such that the expanded theory contains witnesses and is negation faithful. Then the completeness theorem follows immediately.
This paper describes a new approach, the HeKatE methodology, to the design and development of complex rule-based systems for control and decision support. The main paradigm for rule representation, namely, eXtended Tabular Trees (XTT), ensures high density and transparency of visual knowledge representation. Contrary to traditional, flat rule-based systems, the XTT approach is focused on groups of similar rules rather than on single rules. Such groups form decision tables which are connected into...
Within the framework of discrete probabilistic uncertain reasoning a large literature exists justifying the maximum entropy inference process, , as being optimal in the context of a single agent whose subjective probabilistic knowledge base is consistent. In particular Paris and Vencovská completely characterised the inference process by means of an attractive set of axioms which an inference process should satisfy. More recently the second author extended the Paris-Vencovská axiomatic approach...
This paper establishes the equivalence between multilayer feedforward networks and linear combinations of Lukasiewicz propositions. In this sense, multilayer forward networks have a logic interpretation, which should permit to apply logical techniques in the neural networks framework.
In test-based problems, solutions produced by search algorithms are typically assessed using average outcomes of interactions with multiple tests. This aggregation leads to information loss, which can render different solutions apparently indifferent and hinder comparison of search algorithms. In this paper we introduce the performance profile, a generic, domain-independent, multi-criteria performance evaluation method that mitigates this problem by characterizing the performance of a solution by...
A new algorithm for training feedforward multilayer neural networks is proposed. It is based on recursive least squares procedures and U-D factorization, which is a well-known technique in filter theory. It will be shown that due to the U-D factorization method, our algorithm requires fewer computations than the classical RLS applied to feedforward multilayer neural network training.