A novel short-term stock price predicting system

The gray prediction system has been successfully applied to physical control, engineering and economics. The system proposed in this paper can predict the stock event and price instantly at any given time. The authors proposed modified gray prediction systems and compared them with the original and random walk model. Two criteria MSE, MAE were used to compare the performance of the models with Nasdaq Index and S&P 500 daily data. The four models were compared with the random walk model on the basis of the Theil's inequality coefficient. Direction accuracy is also used to measure the models' ability to predict on direction of the evolution of time series. Encouraging results are obtained with these performance measures on the daily stock price both on Nasdaq index and S&P 500 index.