I studied computer science and mathematics at the University of Würzburg and at UT Austin. After that, I finished my doctoral degree with the DMIR Group lead by Andreas Hotho at the L3S Research Center, Hanover and at the University of Würzburg. During this time, I was part of various machine learning and data science projects spanning research fields such as participatory sensing, citizen science, the semantic web, digital humanities and sports sciences.

In this context, I worked on novel methods in the field of exceptional model mining and, for the EU project EveryAware, I pioneered the EveryAware platform. This platform provides a framework for collecting, analyzing and understanding sensor data (for example noise pollution and air quality) in the context of subjective information (for example user perceptions and tags). I am still heavily involved the development process of this platform which enabled follow-up projects like p2map or BigData@Geo.

For my doctoral thesis, I focused on Bayesian hypothesis comparison applied to human navigation behavior. This includes for example task choosing behavior on crowdsourcing platforms, Wikipedia navigation or geo-spatial behavior in cities. For this, I developed various novel methods such as an approach for comparing heterogeneous hypotheses in sequential data or a distributed variant of it’s predecessor method using Apache Spark to handle big data and complex application scenarios. Also, this methodology was the basis for large-scale projects like, e.g., DeepScan.

After having finished my doctoral degree, I am now excited to work in the field of medicine and multiomics as part of Nima Aghaeepour’s lab at Stanford University.

I am interested in methods and concepts from machine learning and data science, including (but not limited to) Bayesian statistics and Bayesian modelling, distributed big data analytics as well as exceptional model mining.

Also, I am particularly intrigued by the interplay of subjective information (e.g., human perceptions and feelings) and objective problem settings. Furthermore, I believe that the usage of background knowledge is an important aspect that I aim to explore especially in the context of small datasets as common in the medical field.

I am passionate about dancing. This includes competing in Latin ballroom, participating in and training Salsa performance teams, as well as some social dancing. Besides that I also like most sports and outdoor activities, reading and, if I get the chance, dabbling in arts.

mgbckr –AT– stanford.edu