Time interval analysis on price prediction in stock market based on general regression neural networks

Abstract

This paper studied the price prediction in sock market based on general regression neural networks (GRNN), and paid more attention to analyzing the time interval effect on the prediction model. Simulations and comparisons were made on actual data and pointed out the relationship between the time interval and the prediction error is not monotonic rise or monotonic decline. Finally, some guidelines for further development were outlined. © 2011 Springer-Verlag Berlin Heidelberg.