A New Hyper-Parameter Optimization Method for Power Load Forecast Based on Recurrent Neural Networks.
Yaru LiYulai ZhangYongping CaiPublished in: Algorithms (2021)
Keyphrases
- recurrent neural networks
- optimization method
- hyperparameters
- chaotic time series
- model selection
- cross validation
- closed form
- bayesian inference
- optimization algorithm
- bayesian framework
- random sampling
- neural network
- em algorithm
- maximum likelihood
- genetic algorithm
- support vector
- optimization process
- noise level
- gaussian process
- optimization methods
- prior information
- differential evolution
- feed forward
- sample size
- particle swarm
- recurrent networks
- maximum a posteriori
- simulated annealing
- short term
- artificial neural networks
- parameter settings
- reservoir computing
- evolutionary algorithm
- incomplete data
- metaheuristic
- incremental learning
- nelder mead simplex
- optimization procedure
- echo state networks
- parameter space
- missing values
- multi objective
- nonlinear dynamic systems
- generative model
- particle swarm optimization
- prior knowledge
- search algorithm
- multiscale
- machine learning