Zdrojový dokument:Scientific papers of the University of Pardubice. Series D, Faculty of Economics and Administration. 12 (2007)
ISSN:1211-555X
Abstrakt:
Managers in their work use sophisticated procedures from many scientific branches. In
order to make good decisions in the stretegic level of management, they can use for
predictions relatively new and effective instruments like neural network algorithms GMDH
and Backpropagation. Their contribution is shown in an example from the transportation
enterprise. There is a time series with 192 values of manipulated transport units per month.
The expert estimations made by the manager working in the transportation enterprise, which
were made on the basis of technological knowledge and experience, were compared with the
estimations made by above mentioned neural network prediction algorithms. In this
comparison the benefit of using neural networks is evident. In the hands of a manager the
neural networks are a powerful tool to increase the benefit of the enterprise.