Topics to work on include (ordered by priority), but are not limited to:
- Theory of Games in Machine Learning: developing new optimization methods, proving convergence, studying Variational Inequalities
- Applying insights of the theory of games in ML, for example, to:
- develop novel multi-player algorithmic approaches for a specific ML problem or area, or novel ML approach for problems in macroeconomics, mechanism design, etc.
- multi-agent reinforcement learning
- solve a specific multi-player application, e.g., competing drones - Robustness in machine learning -- most often formalized as a two-player game, distribution shift
- estimating uncertainty via a multi-player approach
The positions are at CISPA, and the degree is issued by a partner university.
-----------------------------
Please email me (tatjana.chavdarova@berkeley.edu) with your documents: - CV,
- transcript, and
- short motivation letter
when you complete this form.