A Modern Neural Network Model to Do Stock Market Timing on the Basis of the Ancient Investment Technique of Japanese Candlestick

A.M. Kimiagari, M. Jasemi (Iran), and S. Kimiagari (Canada)


Neural network, Technical analysis, Japanese Candlestick, Stock market timing, Finance


This paper presents a new model to do stock market timing on the basis of a supervised feed-forward neural network and the technical analysis of Japanese Candlestick. In this approach the network is not going to learn the candlestick lines alone or in combination, but is to present a kind of regression model whose independent variables are important clues and factors of the technical analysis patterns; and its dependent variable is the market trend in near future. In defining the independent variables two approaches are taken; one is Raw data-based and the other is Signal-based with fifteen and twenty four variables respectively. Experimental results, in which estimated signals are compared with actual events according to real published daily data in Yahoo.finance, showed that the proposed model performs brilliantly well in emission of buy and sell signals while the first approach seems to some extent better then the second.

Important Links:

Go Back