Selective Trimmed Average: A Resilient Federated Learning Algorithm With Deterministic Guarantees on the Optimality Approximation.
Mojtaba KaheniMartina LippiAndrea GasparriMauro FranceschelliPublished in: IEEE Trans. Cybern. (2024)
Keyphrases
- learning algorithm
- approximation guarantees
- relative error
- average cost
- stationary policies
- standard deviation
- approximation algorithms
- monte carlo sampling
- relaxation algorithm
- machine learning
- active learning
- greedy algorithm
- machine learning algorithms
- digital libraries
- reinforcement learning
- training data
- state space
- error tolerance
- classification algorithm
- error bounds
- approximation ratio
- randomized algorithm
- quality guarantees
- linear functions
- wavelet synopses
- data sets
- approximation methods
- approximation error
- markov decision process
- long run
- learning scheme
- closed form
- training examples
- supervised learning
- lower bound
- optimal solution
- neural network