An efficient stock market forecasting model using neural networks

Forecasting financial markets has attracted the interest of neural network researchers. It is a challenging problem, where obtaining a 0.5+e accuracy is an achievement. Researchers applied neural networks successfully to the problems of forecasting currencies, bonds, the futures markets, real estate, and the stock market. In this paper we develop a method for forecasting the stock market. We use novel aspects, in the sense that we base the forecast on fundamental company information, such as earnings per share, price earnings ratio, dividends, sales, profit margin, etc. These indicators and ratios thereof, especially earnings related indicators, are the prime movers of a stock price. The preliminary results we obtain are very promising.