Dominik Alfke

Telefon:

0371-531-+49 371 531 33930

Fax:

+49 371 531 22509

Raum:

Reichenhainer Str. 41, Zimmer 619

E-Mail:

Vita

Born 29.08.1992 in Bassum
Abitur 2011 at Gymnasium Syke
2011 - 2017: Student of Mathematics at Otto-von-Guericke-University Magdeburg; focus on Optimization, Finite Element Methods and Model Order Reduction
Since 01.10.2017 PhD student, Scientific Computing TU Chemnitz

D.Alfke, M. Stoll, Pseudoinverse Graph Convolutional Networks: Fast Filters Tailored for Large Eigengaps of Dense Graphs and Hypergraphs. Data Mining and Knowledge Discovery (2021). Available online at https://doi.org/10.1007/s10618-021-00752-w.

Code is available on GitHub and PapersWithCode.

A previous version was published at https://arxiv.org/abs/1905.10224 under the name Semi-Supervised Classification on Non-Sparse Graphs Using Low-Rank Graph Convolutional Networks.
D. Alfke, D. Potts, M. Stoll, and T. Volkmer, NFFT Meets Krylov Methods: Fast Matrix-Vector Products for the Graph Laplacian of Fully Connected Networks. Frontiers in Applied Mathematics and Statistics, 4 (2018). Available online at https://doi.org/10.3389/fams.2018.00061.

More information and additional material available at the Codes section.