Improved Convergence Rates of Anderson Acceleration for a Large Class of Fixed-Point Iterations.
Casey GarnerGilad LermanTeng ZhangPublished in: CoRR (2023)
Keyphrases
- fixed point
- convergence rate
- stopping criterion
- learning rate
- number of iterations required
- sufficient conditions
- floating point
- policy iteration
- fixed point theorem
- dynamical systems
- constraint databases
- primal dual
- convergence speed
- global convergence
- belief propagation
- machine learning
- linear program
- energy function
- multi objective