Oktober, 2020

Defence of my Master's thesis

Defence of my thesis on the creation of synthetic full-cast audiobooks using deep learning. As part of the thesis, I've developed a demonstrator web application. The thesis and the demonstrator were well received and awarded an excellent grade (1.0, respectively 10/10).

March, 2020

Beautiful Lake Constance

For an machine learning internship at Airbus Defence and Space - Intelligence I moved to Lake Constance. My work focuses on computer vision techniques for fatigue detection, as well as on neural architectures for speech-to-text and text-to-speech applications.

July 26th, 2019

anchorsOnR moves to GitHub

The R-Packet anchorsOnR, which was created during my work as a working student at viadee, is now available on Github! With anchorsOnR one can create high-precision model-agnostic explanations for machine learning models. It allows for highly precise explanations in form of human readable IF-THEN rules, that describe which feature values caused the machine learning model's outcome for one specific observation.

July 9th, 2019

MünsteR (R Users Group)

At the MünsteR R-user group meetup hosted by viadee, I talked about how parallelization with R can be implemented, what preliminary work is necessary and where the challenges and limitations lie.

July 4th, 2019

TPOTR and the Automatic Statistician

For the "Automated Machine Learning (AutoML)" seminar at the University of Münster, we presented our R implementation of the Tree-Based Pipeline Optimization Tool (tpotr) and used the resulting package in our idea of an Automatic Statistician in R (AutoStatR).