I'm a Developer in the area of Data Analysis & Visualization with a focus on Machine Learning and Neural Networks.
I am Information Science student at Regensburg University with focusses on media, digital humanities and machine learning.
Furthermore I work with webservers, websites, databases and android apps - mostly in my spare time.
May 2018 - Oktober 2018
The 8select Software GmbH is a start-up company in the field of e-commerce. They develop software as a service for curated shopping. As a working student in the data department I planed and trained convolutional neural networks for pattern recognition on clothes, which are now used in the productive system at 8select.
Oktober 2018 - Present
In 2018 I continued my studies with the Master's degree Information Science and deepened my knowledge in building bots and other information systems.
In September 2018 I joined a summer school programm at Regensburg's technical university. The programm included basic concepts in machine learning, deep learning and time series analysis. Besides theoretical classes we also had the opportunity to test our new knowledge: we used several tools for data exploration and built neural networks for some tasks, e.g. we trained a small vehicle to drive on a circuit autonomously.
April 2015 - March 2019
The chair of Information Science at Regensburg University focusses on information retrieval, information behavior, interactive assistant systems and computer linguistics. Furthermore i attended courses in media computer science and media science e.g. in human computer interaction, multimedia technologies, chatbot systems, and software programming/engineering. I also attended courses in bioinformatics in which i have learned about analysing DNA-data with NLP methods and machine learning classes.
Here you get an overview over my works as a student and other projects i have done.
Bundestagswahl 2017: Analysis of german politicians' tweet behavior with network analysis and topic modelling.
A self-organizing map is a learning technique with neurons as competitive opponents. At each iteration step a new winner neuron is calculated, which pulls the other neurons with it. In this case this learning tequnique is used to learn a shortest way through geopoints.
Mobile: +49 (0) 174 5258240