Exploring Tunable Hyperparameters for Deep Neural Networks with Industrial ADME Data Sets.
Yadi ZhouSuntara CahyaSteven A. CombsChristos A. NicolaouJi-Bo WangPrashant V. DesaiJie ShenPublished in: J. Chem. Inf. Model. (2019)
Keyphrases
- hyperparameters
- data sets
- neural network
- model selection
- cross validation
- bayesian inference
- closed form
- random sampling
- bayesian framework
- prior information
- support vector
- gaussian process
- maximum a posteriori
- em algorithm
- sample size
- noise level
- maximum likelihood
- gaussian processes
- incremental learning
- training set
- back propagation
- incomplete data
- artificial neural networks
- genetic algorithm
- missing values
- parameter space
- input data
- prior knowledge
- high dimensional data
- high resolution
- parameter settings
- nearest neighbor
- training data
- learning algorithm
- machine learning
- fault diagnosis
- grid search