期刊名称：International Journal of Innovative Research in Computer and Communication Engineering
印刷版ISSN：2320-9798
电子版ISSN：2320-9801
出版年度：2017
卷号：5
期号：10
页码：15777
DOI：10.15680/IJIRCCE.2017.0510017
出版社：S&S Publications
摘要：Predicting stock return is an important financial subject that has attracted researchers’ thought for manyyears. It involves an assumption that fundamental information publicly available in the past has some predictiverelationships in the future stock returns. The predicted value straight affects the stock price and no one take risk to dropdown in stock market index. So due to money involvement and the reputation of the shares, stock market needs to be aperfect or more accurate prediction about their upcoming market trends. Various machine learning algorithms are usedfor stock data set and the objective is to forecast the stock market. In this work the different problems are reviewed,these are defined in the problem formulation and it is resolved with the help of multiple decision tree based learningalgorithms ID3 as previous study C4.5 as proposed method that is implemented in the research work. The results forthe stock market prediction are validated through evaluation metrics, namely mean absolute deviation, mean squareerror, root mean square error, mean absolute percentage error used to estimate the forecasting accuracy in the stockmarket. The proposed model can be a supportive tool for the investors to take the right decision regarding their stocksbased on the analysis of the historical prices of stocks in order to extract any predictive information from that historicaldata. The obtained results show that the proposed C4.5 approach produces better results than the other techniques interms of accuracy.
关键词：Prediction; Stock Market; Data Mining; Prices; Forecast
