A Comprehensive Approach of Algorithmic Trading System based on Technical Analysis

Machine Learning researchers have been attempting to accurately predict stock price trends. Moreover, financial researchers use Technical Analysis with optimization models 1 to achieve the same task by extracting and combining important features for the model. Despite many Algorithmic Trading systems (ATS) available, modelling the stock market has still been proven difficult since stock prices are nonlinear and often appear random 2. This work focuses on leveraging the advantages of both by employing the Technical Analysis approach to create features for training Machine Learning models. The models were evaluated using data from nine stocks listed on the Stock Exchange of Thailand (SET) from 2019 to 2023. The performances of three machine learning models: Logistic Regression, K-Nearest Neighbors, and Support Vector Machines were compared with Differential Evolution and the buy-and-hold strategies. The results showed that the Differential Evolution can significantly improve the return on investment in comparison to the Machine Learning models and the buy-and-hold strategies.