期刊名称：International Journal of Innovative Research in Science, Engineering and Technology
印刷版ISSN：2347-6710
电子版ISSN：2319-8753
出版年度：2017
卷号：6
期号：5
页码：10020
DOI：10.15680/IJIRSET.2017.0605350
出版社：S&S Publications
摘要：This study covers data mining methodology, Stock market parameter set based support vector machinemodel which is proposed to analyze stock market trend direction. In the first part this approach uses various technicalindicators which are useful to discover market trend direction as well as reversal. In the next part stock marketparameters those impose direct or indirect impact on stock market prices are used in support vector machine. These areused as feature vectors selection. Normalization is used to reduce the computation complexity in SVM mathematics.Next part of work is dedicated to sentiment analysis which plays very crucial role to identify impact of news andrelated discussion happening regarding stock prices which is useful in the prediction of stock prices. Prediction abilityof SVM along with technical analysis and sentiment analysis is evaluated in percentages and hit ratio. SVM performsbetter than other forecasting models, proving that the proposed approach is a promising model to predict stock priceprediction.
关键词：Stock trading; Data Mining; Support vector regression; technical indicators; MACD; SMA
