Personalized Projected Ascent for Federated Stackelberg Optimization

Key points

• We therefore study a federated minimax problem where the minimization variable $x$ is shared and aggregated, while each client retains the maximization variable $yi$ (personalized).1
• Because these applications constrain $yi$ to a compact convex set, we analyze projected ascent directly.1
• We show that, in the nonconvex-concave regime, this projected personalized scheme attains the standard sublinear rate.1
• To mitigate this, we propose VR-SGDA, incorporating a variance-reduced ascent step where each client maintains a private variable snapshot.1

Sources (1)

• [1]Personalized Projected Ascent for Federated Stackelberg Optimization
  arXiv (AI, ML, NLP, CV, robotics, multi-agent) · Oct 15, 05:41 AM

  Federated minimax problems, such as distributionally robust optimization and adversarial training, include an inner maximization variable built from each client's local data.

  We therefore study a federated minimax problem where the minimization variable $x$ is shared and aggregated, while each client retains the maximization variable $y_i$ (personalized).

Extractive summary: sentences quoted from the sources.