Verner Vlačić
MMath Cantab. |
Note: Verner Vlačić is no longer with our group.
Publications
- Journal Papers and Manuscripts
-
The extended Smale's 9th problem - on computational barriers and paradoxes in estimation, regularisation, computer-assisted proofs, and learning
A. Bastounis, A. C. Hansen, and V. Vlačić, in preparation. -
Beurling-type density criteria for system identification
C. Aubel, H. Bölcskei, and V. Vlačić, Journal of Fourier Analysis and Applications, Vol. 29, Article No. 45, July 2023. -
Neural network identifiability for a family of sigmoidal nonlinearities
V. Vlačić and H. Bölcskei, Constructive Approximation, May 2021, (invited paper). -
Affine symmetries and neural network identifiability
V. Vlačić and H. Bölcskei, Advances in Mathematics, Vol. 376, Article 107485, pp. 1-72, Jan. 2021.
-
The extended Smale's 9th problem - on computational barriers and paradoxes in estimation, regularisation, computer-assisted proofs, and learning
Supervised Theses
- Diploma Theses
- Studer, Martin, "A Dynamical Systems Perspective on Deep Neural Networks," Spring semester 2020
Supervisor(s): Verner Vlačić - Haymann, Ines, "Complex Analysis and Interpolation for Neural Network Reconstruction," Fall semester 2019
Supervisor(s): Verner Vlačić - Allard, Thomas, "Theoretical Guarantees for Stochastic Gradient Descent," Spring semester 2019
Supervisor(s): Verner Vlačić
- Studer, Martin, "A Dynamical Systems Perspective on Deep Neural Networks," Spring semester 2020
- Student Projects
- Shen, Cheng, "Minima of Optimization Landscapes for Overparametrized Deep Learning Models," Fall semester 2019
Supervisor(s): Verner Vlačić - Allard, Thomas, "Reconstructing a Neural Network from its Output," Fall semester 2018
Supervisor(s): Verner Vlačić
- Shen, Cheng, "Minima of Optimization Landscapes for Overparametrized Deep Learning Models," Fall semester 2019