Abstract: | This article presents the use of neural networks to forecast economic. For economic forecast may be using the following indices: DAX index forecast, forecasting index NIKKEI 225, forecast the index Dow Jones Industrial Average and NASDAQ Composite index forecast. Also, the paper presents a scheme of the implementation process witch supposes seven stages: the selection of the variables, the collection of the dates, the before proceedings dates, the involving, the testing and the validation of the sets, the structure of the network (the presented components are: the number of the input neurons, the hidden layer, the number of output neurons, the accelerated function), criterions of estimating the output network (the main criterions are the number of iterations, the rate of learning). The last part of the paper present the using of the neuronal networks in economic forecasting, particularly the forecasting of Deutsche Aktienindex (DAX) index number. |
