We are happy to announce that Christoph Anders joined our chair.
Before joining our group and writing his thesis at our chair, Christoph worked on a project concerning image segmentation on medical image data to perform tumor segmentation, participated in the development of FIBER, and did various projects as a working student and in student teams.
In his masters thesis "Experimental evaluation of data preprocessing methods for time series classification on brain activity data ", Christoph worked on evaluating the effect data preprocessing can have on EEG data and later task performance regarding time series classification. Excerpt of the abstract: "In EEG data analysis, based on expert knowledge, a lot of effort is spent on data preprocessing utilizing methods used in computational neuroscience, aiming to increase the performance of time-series classification (TSC) models. [...] Despite being identified as a research gap throughout literature, the effect of data preprocessing on later model performance is not fully researched. First, this thesis gives an overview of the methodological background regarding computational neurosciences methods for high-frequency EEG data analysis and how the computer science framework adopted data preprocessing methods. Methods covered in this work range from spatial over spectral to temporal filtering. Second, the data analyzed from Charité, DEAP, and the UCR/UEA TSC archive are presented. More advanced methods and their implications are described in detail. [...] Across DL-models, performance increases from chance-level prediction to close-to-perfect predictions with 95.80% accuracy across subjects of the Charité dataset are shown. [...]"
His general interests span neural networks, deep learning, sports, the language of itself, the representation of language, and in some sense already in the past wearables. We are very interested to see him bridge his research interests regarding neural networks and computational neurosciences with wearables and to follow his academic development. A warm welcome at our chair dear Christoph Anders!
