Stock price forecasting in Forex financial markets by machine learning method

In this article, by using the combination of technical analysis tools and intelligent machine learning method, a method for predicting the stock price trend is presented to give the trader an attitude for buying or selling stocks. In this research, first the required data is collected from the forex market, then among ۲۵ data analysis methods, ten methods are selected with priority according to the dimensionality reduction feature selection method, the output of this step is five intelligent methods of machine learning, linear support vector machine, machine The Gaussian kernel support vector, decision tree, K nearest neighbor and Neobiz method are analyzed using Python programming language. Then, the majority vote method was used for the final decision. Finally, the proposed method has provided a correct prediction rate of approximately ۹۶%. The advantage of the proposed method compared to other analytical methods is that the proposed method has no limitations in using technical analysis methods.

کلیدواژه ها:

technical analysis ، stock price forecasting ، intelligent machine learning methods ، forex ، financial markets