Information Technologies and Neural Network Means for Building the Complex Goal Program “Improving the Management of Intellectual Capital”

Abstract

The article proposes a conceptual approach for finding partial coefficients of influence of alternatives in a complex goal program represented by a linear hierarchy of goals which is considered as a neural network. This becomes possible thanks to the fact that both in the decision tree and in this neural network there are no feedback and single-level connections, as well as threshold sub-goals. As an illustration of this approach, a simplified complex goal program of “Improving the management of intellectual capital of an enterprise” is proposed. To obtain the potential efficiencies of projects in such a CGP the author initially used the DSS “Solon-2”. At the same time, in order to fully automate the proposed process of developing the CGP and its optimization, the author has developed an individual software. The conceptual approach proposed by the author has significant advantages over existing alternative methods. It allows automatically obtaining of the optimal weights of PCI, easily aggregating dynamic changing expert data. The proposed CGP has a great practical interest, since it helps you to optimize the process of managing the intellectual capital of an enterprise, firm or organization. It allows significantly increasing of the profitability of enterprise by obtaining multiple effects such as increasing the productivity of personnel etc.