Učení v Multiagentních systémech za pomocí symbolické reprezentace znalostí

Abstract

This work deals with machine learning with symbolic representation of knowledge using Transparent intensional logic and its application on an element of a multi-agent system. Symbolic methods of machine learning are used as a tool for transforming experiences into formally written knowledge which are used for increasing intelligence of software applications which implements these methods. This work summarizes basics of machine learning, symbolic modes of machine learning, contains theory about Transparent intensional logic and syntax of its computer processable language TIL-Script. In the end analysis and design of an agent implementing symbolic method of machine learning using Transparent intensional logic is presented.