Possibilities of using machine learning in detecting the activities of market makers

This thesis explores the possibilities of using machine learning for detecting the activities of market makers in the foreign exchange market (forex). Market makers are important market participants who provide liquidity by constantly quoting prices for buying and selling currency pairs. However, their activities can also influence and manipulate market prices. The aim of this work is to investigate and evaluate the potential of machine learning for identifying the patterns and strategies that market makers use to influence market prices. The introductory section presents the forex market, the role of market makers, and determines the objectives of the thesis. It then summarizes the current possibilities for detecting market maker activities, including an overview of the principles and types of machine learning. The methodological section describes the selection and configuration of specific machine learning algorithms, the integration of data from the MetaTrader 4 trading platform using Node.js and TensorFlow.js. The process of obtaining, formatting and processing data is described. Furthermore, procedures are proposed for testing and verifying the created models. In the discussion, the results are compared with the set objectives and possibilities for practical use of the results are proposed, including the identification of areas for further research. The conclusion summarizes the main findings of the work and the fulfillment of the set objectives.