Bayesian discrete choice modeling

I am interested in analyzing discrete choice behavior, especially preference heterogeneity. I am currently implementing the R package RprobitB for Bayesian estimation of latent class probit models.

Initializing numerical optimization

I am interested in methods for efficient initialization of numerical optimization. I recently started the ino project together with Marius Oetting for providing initialization tools.

Financial hidden Markov models

I am interested in modeling financial data with hidden Markov models (HMMs) - a versatile class of statistical models for time series data that is assumed to be dependent on latent states. In case of financial data, the latent states can be interpreted as different moods of the market. Even though these moods cannot be observed directly, price changes - which clearly depend on the current mood of the market - can be observed. Thereby, using an underlying Markov process, we can detect which mood is active at any point in time and how the different moods alternate. Depending on the current mood, a price change is generated by a different distribution. These distributions characterize the moods in terms of expected return and volatility.

In my master thesis, I applied the hierarchical extension of HMMs to German stock index data, aiming to improve the model’s capability for distinguishing between short- and long-term trends and interpreting market dynamics at multiple time scales. Together with Timo Adam, I published the work in the Journal of Statistical Modelling and implemented the method in the R Package fHMM.