Terry Lyons, Harald Oberhauser and Hao Ni

Multidimensional sequential data appears both in time series analysis, representing a process evolving in time, and in shape analysis, representating the boundary of an object. Extracting geometric features of, and detecting symmetries in, such data is useful in their mathematical analysis and in improving the efficiency of machine learning algorithms. In recent years, both the iterated-integral signature and the moving frame method have been successfully used in this endeavor. They both provide geometrically relevant representations of sequential data, and have been used in applications ranging from mathematical finance to computer vision.

The aim of this workshop is to showcase the recent applications of these mathematical tools in data science. Introductory talks on the iterated-integrals signature and the moving frame method will provide a common language for the research presentations on current applications.

Plenary speakers are:

  • Irina Kogan (NC State University)
  • Terry Lyons (University of Oxford)
  • Harald Oberhauser (University of Oxford)
  • Peter Olver (University of Minnesota)

Confirmed speakers are:

  • Mireille Boutin (Purdue University)
  • Kurusch Ebrahimi-Fard (Norwegian University of Science and Technology)
  • Adeline Fermanian (LPSM, Sorbonne Université)
  • Eric Geiger (NC State University)
  • Stephan Huckemann (Universtiy of Göttingen)
  • Shannon McPherron (Max Planck Institute for Evolutionary Anthropology)
  • Simon Neubauer (Max Planck Institute for Evolutionary Anthropology)
  • Hao Ni (University College London)
  • Rosa Preiß (Technical University Berlin)
  • Nikolas Tapia (Weierstrass Institute for Applied Analysis and Stochastics Berlin)
  • Christoph von Tycowicz (Zuse Institute Berlin)
  • Katrina Yezzi-Woodley (University of Minnesota)