Table 1 Summary of H2O-AutoML models.

DRF

Parallel processing, ensemble of trees

Scalable, robust predictions^56,39

GBM

Boosting, loss function optimization

Handles complex tasks, iterative improvement^56,39

GLM

Link function, multiple distributions

Flexible for various predictions^56,39

NN

Multilayer perceptron, gradient descent

Accurate, handles non-linear relationships³⁹

XRT

Random feature thresholds

Reduces bias, learns non-linear models³⁹

Stacked ensemble learning

Combines algorithms

Improves model performance^56,39