Compare Hidden Markov Model And Artificial Neural Network For Predicting Stock Price

Abstract

In recent years, predicting the stock’s future price based on the past data has never ceased to draw researchers and investors’ attention. Because of the high fluctuation of the stock market every day, stock price forecasting is a complicated issue to get accurate data to compare to actual data. There are many approaches were used to solve this problem such as ARIMA model, Hidden Markov models (HMM), Artificial Neural Networks (ANN),... But each model has its own limitation, there are some problems occur like data of financial time series are nonlinear or predicted results have large errors compare to actual values,...In this paper, we compare two single models: Hidden Markov models (HMM) and Artificial Neural Networks (ANN) for predicting the price of the stock in Ho Chi Minh stock exchange. First, we give a description of each model. Next, we apply this model to real data in order to forecast the data and information about the stock in the future. Finally, based on the results given of two models, we will compare them and determine the better model perform well in predictions.