Decentralized personalized federated learning: Lower bounds and optimal algorithm for all personalization modes.
Abdurakhmon SadievEkaterina BorodichAleksandr BeznosikovDarina DvinskikhSaveliy ChezhegovRachael TappendenMartin TakácAlexander V. GasnikovPublished in: EURO J. Comput. Optim. (2022)
Keyphrases
- learning algorithm
- worst case
- dynamic programming
- lower bound
- objective function
- optimal solution
- computational complexity
- np hard
- lower and upper bounds
- learning problems
- competitive ratio
- theoretical analysis
- constant factor
- probabilistic model
- cost function
- active learning
- learning process
- search space
- optimal cost
- error bounds
- combinatorial optimization
- detection algorithm
- linear programming
- reinforcement learning