Newer
Older
# Diploma Thesis
Min-max optimization methods inadversarial learning - Robust learning of Neural Network models, including S4at FEL ČVUT / FEE CTU in Prague, bc. Tomáš Kasl
### The thesis document
can be seen [Here:](https://gitlab.fel.cvut.cz/kasltoma/diploma-thesis/-/blob/main/thesis.pdf)
### Final presentation:
can be seen [Here:](TBD)
### The robust training can be run,:
in [main.py:](https://gitlab.fel.cvut.cz/kasltoma/diploma-thesis/-/blob/main/main.py)
or in [compare_models_on_robustness.py]
Based on [annotated S4](https://srush.github.io/annotated-s4/) and [solving optimization problem](https://github.com/optimization-for-data-driven-science/Robust-NN-Training)