Fundamental Analysis of Stock Trading Systems using Classification Techniques

The traditional forecasting of revenue growth rate (RGR) is based on normal distribution. Due to emergence of information technology today, data mining has become one of important research trends. Therefore, this paper mainly forecasts revenue growth rate of firms in stock trading systems by classification techniques. It is very important instrument for investors that correctly predict future growing firms from data of fundamental analysis in trading systems, because the accurate prediction of RGR will bring huge profit for investors in the future. This paper proposes a process to predict RGR of firms, which employs Decision tree C4.5, Bayes net, Multilayer perceptron and Rough sets techniques. Moreover, the paper uses the actual RGR dataset in Taiwan stock market to illustrate the proposed process. From the results, we recommend the rough set as analysis tool because the performance is superior to the listing methods and understandable rules are produced.