Research on Cup and Handle Price Pattern Based on Svm Method

Abstract

Analysis of the data of stock is one of the most important methods for investor to knowing if the stock is going to have a uptrend and determine whether call and put the stock. This is one of the methods that can highly decrease the risk of investing. Therefore, to decrease the risk of loss and increase the return, prediction of the stock price become a crucial task and has attracted many investors. To introduce a new method to solve the problem, this paper didn’t focus on predicting the price of stock but cup and handle pattern. This pattern always implies that there is an upward trend follows this. The trouble is how to find out the stock has the pattern from many stock data. Based on machine learning (ML) techniques, this paper uses support vector machine (SVM) model to classify the data and to determine whether this set has the pattern or not. The model uses gaussian radial function as kernel function. The experiment result demonstrates that the classify model reach a relatively satisfied result and can give a relative good, classified result.