In this project we used two Machine Learning algorithms namely, Recurrent Neural Networks and Reinforcement Learning to support an algorithmic trading strategy based on high-frequency financial data. Such data came from the NYSE Trade and Quote (TAQ) database which contains intraday transactions for all securities listed on the New York Stock Exchange (NYSE) and American Stock Exchange (AMEX), as well as the Nasdaq National Market System (NMS).