Advanced Financial Data Processing and Labeling Methods for Machine Learning

Abstract

The use of machine learning in the prediction of financial asset prices is gaining interest in both academic research and the financial industry. Although these algorithms have been successful in many applications, predicting stock market movements is still a real challenge for machine learning. This work investigates the potential benefit of advanced financial data preprocessing methods prior to the application of machine learning. The results of the developed model confirm the relevance of these methods and their ability to address some of the limitations that characterize the research work on price prediction by machine learning. The integration of these methods into machine learning prediction models brings them in line with the reality of the financial market and therefore allows for better investment decisions.