Which hyperparameters to optimise?: an investigation of evolutionary hyperparameter optimisation in graph neural network for molecular property prediction.
Yingfang YuanWenjun WangWei PangPublished in: GECCO Companion (2021)
Keyphrases
- hyperparameters
- neural network
- model selection
- cross validation
- genetic algorithm
- closed form
- bayesian inference
- prediction model
- bayesian framework
- support vector
- random sampling
- gaussian process
- gaussian processes
- sample size
- prior information
- em algorithm
- maximum a posteriori
- maximum likelihood
- noise level
- incremental learning
- prediction accuracy
- parameter optimization
- incomplete data
- missing values
- back propagation
- grid search
- artificial neural networks
- regression model
- multiscale
- parameter space
- bp neural network
- energy function
- fitness function
- markov random field
- training set
- radial basis function