Bayesian Clustering and Variance-Covariance Matrix Modelling using the Dirichlet Process Prior

In Short

To conclude my research master Economic and Financial Research: Track Econometrics, I wrote my master thesis on the applicability of Bayesian Clustering models to model Variance-Covariance matrices over time with the goal to model and predict chaotic and seemingly unrelated processes such as a group of stocks. All in all I’ve spent almost a year, with some pauses to follow courses, on the thesis worth in total 30 ECTS. Although I have really enjoyed some parts of that process, others really made me reflect on the direction I would want to start my carreer after my master.

Research Master Background

Even more than a standard master, the research master prepares students for a carreer in academia. There is more attention for the conduction, communication and presentation of your research. Next to that, you have access to more advanced courses that are really only given to research master students as well as the regular econometric master courses. As such, I was able to enjoy the courses: Risk Management, Financial Markets, High-Dimensional Econometric Methods for Big Data, Advanced Operations Research (following Project Euler) and Advanced Concepts in Machine Learning.

The thesis

In the end, I did not see myself work that on very academic, quite theoretical problems and thus for a carreer outside of academia.