Students:
- Daniil Tiapkin
- Sultan Bakiev
- Ekaterina Solodneva
- Nikita Vailiev
Idea: use Eigengame algorithm [1,2] to compute (truncated) SVD of large-scale matrices and compare it to classical implementations of SVD (such as randomized SVD [3]).
- [1] Gemp et al. EigenGame: PCA as a Nash Equilibrium. Ref: https://arxiv.org/abs/2010.00554
- [2] Gemp et al. EigenGame Unloaded: When playing games is better than optimizing Ref: https://arxiv.org/abs/2102.04152
- [3] Sklearn implementation of RandomSVD. Ref: https://scikit-learn.org/stable/modules/generated/sklearn.utils.extmath.randomized_svd.html