Stock Trading Using Machine Learning

This project presents an approach to maximally increase profits for stock trading. Both Fundamental and Technical analysis are used in the portrayed approach. A large number of companies from various sectors were considered, and sector wise grouping of these companies is done. First, the best companies to invest in each sector are found by doing a fundamental analysis of the stock. Next, portfolio optimization is done to figure out how much cash should be allocated to each company at any amount of time. Once this is done, the technical factors are used to figure out buying and selling points, and set up a daily routine to optimize the current price in hand. This can either be implemented using a branching model or by various Machine Learning models such as Reinforcement Learning or Neural Networks. Each of these models is used and the trader is encouraged to rely on the one which gives the best results according to the comparisons made in the thesis.