The stock market is the backbone of the financial status of any country. The unpredictable behaviors of the stock market change the mental group of investors and buyers. If the stocks traders predict right trend in stock price, they can realize profits. Therefore, prediction of stock price is very important factor for buyers and seller in stock market. The accuracy and behavior is accurate in limited data of stocks, if the range of data are increases the impact of prediction is decrease. The recent research in the area of stock price prediction with machine learning methodologies makes use mainly of SVM architectures and takes into account several predictors. The aim of this proposed work is to increase the capacity of sample size of classifier and increase the accuracy of prediction with the derivation of cascading and ensemble classifier.