Given the significance of stock prices in most companies' investment strategies, accurately forecasting them amidst high volatility is crucial. Stock markets are renowned for their complex, turbulent, and opaque nature within the global financial landscape. Researchers have consistently sought to develop methods for predicting stock prices with greater accuracy and lower error rates. Market forecasting is primarily undertaken using technical and fundamental analysis. Technical analysis involves evaluating and forecasting market price trends using numerous indicators available on online platforms provided by exchanges or brokers. The abundance of these indicators often leads to confusion and challenges in selecting the most suitable ones for technical analysis. This research aims to identify the most effective technical analysis indicators within the Forex market and to compare the performance of various machine learning methods in predicting price trends in the Forex market. To this extent, daily data for two major currency pairs (EUR/USD and USD/JPY) and two minor pairs (CAD/JPY and EUR/CHF) were collected over 15 months. Thirty commonly used technical indicators were selected and ranked using the ReliefF feature selection method to identify the top ten indicators. The outputs from this stage were then modeled using six machine learning algorithms: artificial neural networks, decision trees, random forests, support vector machines, K-nearest neighbors, linear regression, and logistic regression, implemented in MATLAB. The analysis results indicate that the artificial neural network outperformed the other machine learning methods, achieving the lowest MAPE and RMSE and providing the most accurate predictions.
Saeidi Shahram (Thu,) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: