Elham Ahmadi - M.S. Student, Department of Industrial Engineering, Yazd University, Yazd, Iran
Mohammad HosseinAbooie - Ph.D, Assistant Professor of Industrial Engineering,YazdUni, Yazd, Iran

In this paper, the nonlinear autoregressive model with exogenous variables as a new neural network is used for timing of the stock markets on the basis of the technical analysis of Japanese Candlestick. In the model the nonlinear autoregressive model with exogenous variables” is an analyzer. For a more reliable comparison, here (like the litrature) two approaches of Raw-based and Signal-based are devised to generate the input data of the model. The correct predictions percentages for periods of 1- 6 days with the total number of buy and sell signals are considered. The result prove that to some extent the approaches have similar performances while apparently they are superior to a feed-forward static neural network. The created network is evaluated by the measure of Mean of Squared Error and the proposed model accuracy is calculated to be extremely high.

Finance; Stock Market Forecasting; Technical Analysis; NARX recurrent neural network; Levenberg Marquardt algorithm