Theoretical analysis of Adam using hyperparameters close to one without Lipschitz smoothness.
Hideaki IidukaPublished in: Numer. Algorithms (2024)
Keyphrases
- hyperparameters
- theoretical analysis
- prior information
- model selection
- cross validation
- bayesian inference
- bayesian framework
- support vector
- closed form
- em algorithm
- gaussian process
- random sampling
- sample size
- noise level
- maximum likelihood
- incremental learning
- maximum a posteriori
- regularization parameter
- incomplete data
- parameter space
- gaussian processes
- missing values
- prior knowledge
- higher order
- parameter settings
- expectation maximization
- genetic algorithm
- input data