DIKU Bits: Data Analysis in Curved Spaces
Stefan Horst Sommer, Professor in the Image Analysis, Computational Modelling and Geometry section at the Department of Computer Science.
In the applied geometry lab in the IMAGE section, we use methods from differential geometry to analyze data with geometric structure. This can for example be observations on the surface of the earth (which is curved) or the shape of human organs influenced by progressing disease. In the talk, I will give examples of where geometry appears in machine learning, statistics and data science, and talk about the research using geometric methods we are currently pursuing
Zooming in on Stefan Horst Sommer
Which courses do you teach? (BSc and MSc)
I am teaching the course "Artificial Intelligence" and besides that, I am head of studies for the Bachelor's programme in Machine Learning and Data Science.
Which technology/research/projects/startup are you excited to see the evolution of?
The development of artificial intelligence, especially when very large amounts of data are used, thereby the algorithms have the potential to become much better than humans at performing tasks. Examples can be language processing like the GPT-3, or self-driving cars that possibly can be trained to completely eliminate the concept of road accidents.
What is your favorite sketch from the DIKUrevy?
Unfortunately, I have not yet watched the DIKUrevy.