Proximal Point

Proximal point methods are iterative optimization algorithms that solve complex problems by iteratively approximating a simpler subproblem. Current research focuses on extending these methods to handle large-scale problems, particularly in federated learning and mean field games, often employing variants like Halpern-type or Bregman proximal point algorithms, and unfolded proximal neural networks. This work is significant because it improves the efficiency and robustness of optimization across diverse fields, including image processing, machine learning, and game theory, leading to more accurate and scalable solutions.