A new hybrid approach based on k-means and linear regression to predict stock price and direction in short-term period

Stock price and direction prediction plays an important role in investor’s decision making process. The non-linear behavior of stock market made the stock price prediction a challenging problem for researchers. In this study, we propose a hybrid model based on k-means clustering algorithm and linear regression, in order to predict stock price and direction in short-term period. In spite of the fact that stock market does not put on a happy face at predictions, it has been represented that proposed model works well and is superior to the linear regression alone

Keywords:

stock price prediction , k-means , linear regression