Whether analysing classification algorithms for academic research or optimising ERP workflows for a mid-sized company, my goal remains the same: solving complex problems through data.

The following is a collection of my work in Econometrics, Machine Learning, and Business Intelligence.

This thesis compares Linear Discriminant Analysis, Quadratic Discriminant Analysis, and Random Forests to evaluate their suitability for binary classification tasks, with a particular focus on predicting diabetes status from physiological and demographic data.
Motivated by the principle that no single model performs best across all settings, the study constructs three artificial datasets that each reflect different structural assumptions, enabling a controlled assessment of model behaviour when assumptions are met or violated. The methods are further applied to the real‑world PimaIndiansDiabetes2 dataset.

Proseminar: `School Choice with Consent´

Group presentation from the seminar on scientific work

Chapters 2.2 Kidney Exchange Problem, 2.3 (TTCC) Kidney Exchange Mechanism and  Chapter 3.3 Own Example prepared and presented