The Hidden Markov model (HMM) is a probabilistic model that can be used to predict economic regimes and stock prices. In this project the students will gather some type of economic data (say the performance of a particular stock) and try to model it with a HMM. In particular, one needs to tune the different parameters of the model over a training subset of the data and then see how good the approximation is on the unseen values. Apart from learning about the mathematical structure of the HMM, the students will explore issues concerning model selection and evaluation of performance.
Schedule: To be determined with the student(s).
Supervisors: George Kerchev
Difficulty level: Introductory/intermediate (some basic knowledge of probability theory is needed)
Tools: Programming can be done in C/C++, Python, R, Matlab/Octave, Wolfram Mathematica, Maple etc., depending on the knowledge of the student.
Results: To be completed at the end of the project.