Analytics and Statistics

Examples of Bachelor and Master Theses

As an example, we show a selection of theses that have already been completed at the chair. For all work, we require relevant statistical knowledge that has been acquired in the context of our courses and seminars, as well as very good programming and mathematical knowledge. Finding a suitable topic is always an individual process. If you are interested, please contact the secretariat with your current performance overview and a short résumé.

  • Development of a supply demand model for the Italian gas market.
  • Displacement interpolation.

  • Principal component analysis and its use in monitored machine learning. 

  • Smoothed Bootstrap: An Investigation of It's Effect on the Dependence Structure. 

  • Multivariate self-exciting point processes and applications. 

  • Variable selection and handling of metric features in decision trees. 

  • Development of a forecast model to predict the amount of baggage transferred per flight at Frankfurt Airport. 

  • Optimal forecasting under asymmetric loss - a Bayesian approach-.

  • Correspondence analysis on data from copula families. 

  • Application of machine learning methods for production process optimization. 

  • Time Series Compression for Classification of Vibration Signals.

  • Machine Learning in Medicine: Can Decision Trees provide a comprehensive and powerful improvement of in-hospial mortality prediction of ICU patients? 

  • Copula-based forecast aggregation. 

  • The simplifying assumption in the pair-copula construction. 

  • Dependence Modeling for Operational Risks. 

  • Active Power Forecasting using ARIMA Model and Neural Network Approaches. 

  • Stability of the estimate. 

  • Gluing Copulas: Simulation and Semi-Parametric Estimation. 

  • Stylized Facts of Macroeconomic and Wind Power Forecast Errors.

  • The Local Outlier Factor - An algorithm for analyzing outliers. 

  • Polytome Logit and Probit Regression.

  • Monaural localization on the horizontal plane using HRTF. 

  • Multivariate analysis methods in automatic face recognition.