Forecasting short-term stock price indexes - an integrated predictor vs. neural network predictor

Neural network predictor is widely applicable in intelligent forecast. However, it encounters the crucial problem that the predicted values always cannot achieve the satisfactory results because the generalization capability in neural network predictor can´t perform extrapolation well. Therefore, this study introduced an integrated predictor to compare with neural network predictor utilized for the applications of non-periodic short-term time series forecast. This proposed predictor in fact is an integrated model, combing a grey prediction model and a cumulative least squared linear prediction model, with the technique of automatically compensating a possible overestimated predicted value by a potential damped predicted value around those predicting points having locally extreme high or low value. The verification of this study also experiments successfully in the stock price indexes forecast, and the results out of integrated predictor achieved the best accuracy on the predicted stock price indexes compared with Back-Propagation neural network predictor, Box-Jenkins, and Holt-Winters smoothing.