SVM, a supervised machine learning algorithm, can be used for both regression and classification problems. Moreover, SVM shows that it can perform well on non-linear dataset problems, based on the kernel we choose in training SVM model. SVM have been widely used for stock market prediction.