Stochastic distributed learning with gradient quantization and double-variance reduction.
Samuel HorváthDmitry KovalevKonstantin MishchenkoPeter RichtárikSebastian U. StichPublished in: Optim. Methods Softw. (2023)
Keyphrases
- distributed learning
- variance reduction
- gradient estimation
- monte carlo
- policy gradient
- sample size
- importance sampling
- collaborative learning
- quasi monte carlo
- markov chain
- bias variance decomposition
- naive bayes classifier
- knowledge integration
- confidence intervals
- particle filter
- distributed learning environments
- text mining
- training data