Projects funded by the NCN
EAGER: IMPRESS-U: Random Matrix Theory and its Applications to Deep Learning
2023/05/Y/ST1/00188
Keywords:
random matrices deep neural networks input-output Jacobian Fisher information matrix Hessian neural tangent kernel empirical spectral distribution central limit theorem concentration of measure singular values
Descriptors:
- ST1_13:
- ST1_16:
Panel:
ST1 - Mathematics: all areas of mathematics, pure and applied, as well as mathematical foundations of computer science, physics and statistics
Principal investigator (from the host institution):
Number of co-investigators in the project:
Call: IMPRESS-U
Amount awarded: 358 802 PLN
Project start date (Y-m-d): 2024-03-26
Project end date (Y-m-d): 2026-03-25
Project duration:: 24 months (the same as in the proposal)
Project status: Pending project