
Informed meta-learning is a novel paradigm that aims to develop domain-agnostic meta-learners by integrating external knowledge as an additional source of inductive biases, complementing the purely data-driven approach of conventional meta-learning.


coremsg

informed-meta-learning-enhancing-data-efficiency-and-robustness-by-integrating-expert-knowledge-into-adaptive-learning-algorithms


Informed Meta-Learning: Enhancing Data Efficiency and Robustness by Integrating Expert Knowledge into Adaptive Learning Algorithms
