Clustering-Classification Based Prediction of Stock Market Future Prediction

Abstract — Stock market values keeps on changing day by day, so it is very difficult to predict the future value of the market. Although there are various techniques implemented for the prediction of stock market values, but the predicted values are not very accurate and error rate is more. Hence an efficient technique is implemented for the prediction of the stock market values using hybrid combinatorial method of clustering and classification. The dataset is taken from shanghai stock exchange market and is first clustered using K-means clustering algorithm and these clustered values are classified using horizontal partition decision tree. Keywords—Stock market, K-mean, Prediction. I