TITLE:
A Neuron Model with Dendritic Nonlinearity for Predicting the Influence of Overreaction in Shanghai Stock Market
AUTHORS:
Sha Zijun, Hu Lin
KEYWORDS:
Neuron Model with Dendritic Nonlinearity, Overreaction, Prediction, Stock Market
JOURNAL NAME:
Chinese Studies,
Vol.4 No.1,
January
23,
2015
ABSTRACT: On the basis of previous research, this study assembled trading data from January 2004 to October 2014 to verify the overreaction in Shanghai stock market. The influence of overreaction decreases with time from 2007 onwards and turns to disappear from 2011. The neuron model with dendritic nonlinearity (NMDN) proposed in this paper fits and predicts the variability of abnormal returns of ill-preforming and well-preforming stocks in the test period. This experiment demonstrates that NMDN possesses high computational ability and succeeds to predict trends in the influence of overreaction.