Surrogate-Assisted Hybrid-Model Estimation of Distribution Algorithm for Mixed-Variable Hyperparameters Optimization in Convolutional Neural Networks.
Jian-Yu LiZhi-Hui ZhanJin XuSam KwongJun ZhangPublished in: IEEE Trans. Neural Networks Learn. Syst. (2023)
Keyphrases
- hybrid model
- estimation of distribution algorithms
- hyperparameters
- convolutional neural networks
- feature subset selection
- cross validation
- particle swarm optimization algorithm
- model selection
- evolutionary computation
- support vector
- multi objective
- artificial neural networks
- bayesian framework
- support vector regression
- closed form
- bayesian inference
- evolutionary algorithm
- combinatorial optimization
- random sampling
- incremental learning
- em algorithm
- particle swarm optimization
- genetic algorithm
- support vector machine svm
- noise level
- prior information
- incomplete data
- multi objective optimization
- simulated annealing
- combinatorial optimization problems
- maximum a posteriori
- sample size
- maximum likelihood
- genetic programming
- optimization problems
- optimization algorithm
- support vector machine
- evolution strategy
- genetic algorithm ga
- fuzzy logic
- prior knowledge
- image segmentation
- computer vision