Meta-learning the learning trends shared across tasks

J Rajasegaran, S Khan, M Hayat, FS Khan… - arxiv preprint arxiv …, 2020 - arxiv.org
Meta-learning stands for'learning to learn'such that generalization to new tasks is achieved.
Among these methods, Gradient-based meta-learning algorithms are a specific sub-class …

Meta-learning with adjoint methods

S Li, Z Wang, A Narayan, R Kirby… - … Conference on Artificial …, 2023 - proceedings.mlr.press
Abstract Model Agnostic Meta-Learning (MAML) is widely used to find a good initialization
for a family of tasks. Despite its success, a critical challenge in MAML is to calculate the …

[PDF][PDF] Meta-Learning with Adjoint Methods

Z Wang, A Narayan, RM Kirby, S Zhe - 2023 - par.nsf.gov
Abstract Model Agnostic Meta Learning (MAML) is widely used to find a good initialization
for a family of tasks. Despite its success, a critical challenge in MAML is to calculate the …

A Markov decision process approach to active meta learning

B Wang, A Koppel, V Krishnamurthy - arxiv preprint arxiv:2009.04950, 2020 - arxiv.org
In supervised learning, we fit a single statistical model to a given data set, assuming that the
data is associated with a singular task, which yields well-tuned models for specific use, but …