Module meta_learning_kernels 

Meta-Learning for Kernel Selection

This module implements meta-learning strategies for automated kernel selection, few-shot kernel learning, transfer learning for kernels, and neural architecture search for kernel methods.

References

• Vanschoren et al. (2014): “Meta-Learning: A Survey”
• Feurer & Hutter (2019): “Hyperparameter Optimization”
• Hospedales et al. (2021): “Meta-Learning in Neural Networks: A Survey”
• Wilson & Izmailov (2020): “Bayesian Deep Learning and a Probabilistic Perspective of Generalization”

Structs

DatasetMetaFeatures

Meta-features extracted from datasets for kernel selection

MetaLearningConfig

Configuration for meta-learning kernel selection

MetaLearningKernelSelector

Meta-Learning Kernel Selector

TaskMetadata

Task metadata for meta-learning

Enums

MetaKernelType

Kernel types supported by meta-learning

MetaLearningStrategy

Strategy for meta-learning-based kernel selection

PerformanceMetric

Performance metrics for kernel evaluation