We use Neural Networks algorithms for forecasting financial time series. We check first he kind of correlations that the series exhibits by means of the estimate of the Hurst's H parameter. The range of correlations characterize the time series and fives useful hints for choosing the network and the training set. The rime series ocnsidered is given by the values of the futures of Italian BTP.
Neural Networks for financial forecast / Rotundo, Giulia; Tirozzi, Benedetto; Valente, M.. - STAMPA. - (1998), pp. 351-356. (Intervento presentato al convegno ESANN '98 6-th European Symposium on Artificial Neural Networks tenutosi a Bruges, Belgium nel April 22-24, 1998).
Neural Networks for financial forecast
ROTUNDO, GiuliaInvestigation
;TIROZZI, BenedettoSupervision
;
1998
Abstract
We use Neural Networks algorithms for forecasting financial time series. We check first he kind of correlations that the series exhibits by means of the estimate of the Hurst's H parameter. The range of correlations characterize the time series and fives useful hints for choosing the network and the training set. The rime series ocnsidered is given by the values of the futures of Italian BTP.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
