What type of study is this?

Nesterov Accelerated Gradient Tracking With Adam for Distributed Online Optimization

View Full Paper

Ask AI

Bookmark

The NGTAdam algorithm shows improved convergence performance in distributed online optimization problems.
Its dynamic regret is assessed through linear system inequality and is influenced by initial conditions.
Numerical experiments demonstrate that NGTAdam outperforms other distributed online optimization algorithms.
If the upper bound on dynamic regret is sublinear, the overall regret shows similar behavior over time.

No takes yet. Share an insight, caveat, or question.

This algorithm combines nesterov acceleration and adaptive moment estimation, improving dynamic regret in distributed online optimization.

Su et al. (2025) studied this question.

View Full Paper

Ask AI

Bookmark

This algorithm combines nesterov acceleration and adaptive moment estimation, improving dynamic regret in distributed online optimization.