Artificial intelligence

Abstract: Despite of impressive successes of one of most actively developing and perspective sciences of the modernity which has obtained a title – the artificial intelligence, the extremely complicated area in it still is formalising of some the tasks solved by the person. There are numerous attempts of its simplification on the basis of fuzzy logic, artificial neural nets and other similar methods. As a result of these researches the complicated designer from the heterogeneous components concerning to various scientific directions and created under each specific target turns out. One of base directions of overcoming of existing difficulties is construction of adaptive, hierarchical architectures. However this direction frightens off many talented researches speaking about exponential increasing complexity of designing of a system. In the given article it is offered enough the simple formalism, permitting to build hierarchical architectures on the basis of artificial neural nets.

Keywords:

References: