In this thesis project, we want to combine the best of two worlds:
Forecasting stock prices is a well-developed task in machine learning and so is control theory as basis for automatic decision making.
Optimal trading control is a natural marriage between a forecasting task and a control task. Here, you will develop a highly multi-disciplinary approach to stock trading.
You tackle a whole set of challenges: What tools may we use to influence the outcome of our portfolio (e.g. stocks, bonds, options, annuities)? How can we make a good forecast of the relevant stock prices? What kind of trading strategy should we employ and how can we optimize it?