Epsilon Greedy Strategy for Hyper Parameters Tuning of A Neural Network Equalizer.
Quyet NguyenNoel TekuTamal BosePublished in: ISPA (2021)
Keyphrases
- hyperparameters
- greedy strategy
- neural network
- model selection
- greedy algorithm
- cross validation
- parameter settings
- closed form
- bayesian inference
- random sampling
- support vector
- bayesian framework
- prior information
- em algorithm
- noise level
- maximum a posteriori
- variational bayes
- posterior distribution
- incremental learning
- sample size
- maximum likelihood
- artificial neural networks
- bayesian methods
- missing values
- dynamic programming
- incomplete data
- conjugate priors
- parameter values
- noisy images
- markov random field
- higher order
- parameter space
- upper bound
- active learning
- prior knowledge
- search space
- reinforcement learning
- image processing