Greifswald University Medical Centre, a leading university hospital, is home to the Institute of Bioinformatics, where research focuses on biometrics and high-dimensional statistics, mathematical modeling of important biomedical molecular processes, and bioinformatics and integrative data analysis using state-of-the-art machine learning techniques. I have been involved in an interdisciplinary research project with the Department of Trauma Surgery at the University of Greifswald on tissue type classification based on Raman spectroscopy data. Using the programming language R, I applied different pattern recognition methods (e.g. support vector machines, decision trees, etc.) to the provided data and evaluated the results. I then presented the results to our collaborator in a presentation, thereby supporting their research project by providing an approach that allowed for automatic classification of the tissue type at hand.
Under guidance of Prof. Dr. rer. nat. Lars Kaderali.
Example specturm in the data set:
The spectra (dots) and their classes (circles) plotted across the two largest principal components:
