Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networks
C. Finn, P. Abbeel, and S. Levine. (2017)cite arxiv:1703.03400Comment: ICML 2017. Code at https://github.com/cbfinn/maml, Videos of RL results at https://sites.google.com/view/maml, Blog post at http://bair.berkeley.edu/blog/2017/07/18/learning-to-learn/.
Abstract
We propose an algorithm for meta-learning that is model-agnostic, in the
sense that it is compatible with any model trained with gradient descent and
applicable to a variety of different learning problems, including
classification, regression, and reinforcement learning. The goal of
meta-learning is to train a model on a variety of learning tasks, such that it
can solve new learning tasks using only a small number of training samples. In
our approach, the parameters of the model are explicitly trained such that a
small number of gradient steps with a small amount of training data from a new
task will produce good generalization performance on that task. In effect, our
method trains the model to be easy to fine-tune. We demonstrate that this
approach leads to state-of-the-art performance on two few-shot image
classification benchmarks, produces good results on few-shot regression, and
accelerates fine-tuning for policy gradient reinforcement learning with neural
network policies.
Description
[1703.03400] Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networks
cite arxiv:1703.03400Comment: ICML 2017. Code at https://github.com/cbfinn/maml, Videos of RL results at https://sites.google.com/view/maml, Blog post at http://bair.berkeley.edu/blog/2017/07/18/learning-to-learn/
%0 Journal Article
%1 finn2017modelagnostic
%A Finn, Chelsea
%A Abbeel, Pieter
%A Levine, Sergey
%D 2017
%K meta-learning
%T Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networks
%U http://arxiv.org/abs/1703.03400
%X We propose an algorithm for meta-learning that is model-agnostic, in the
sense that it is compatible with any model trained with gradient descent and
applicable to a variety of different learning problems, including
classification, regression, and reinforcement learning. The goal of
meta-learning is to train a model on a variety of learning tasks, such that it
can solve new learning tasks using only a small number of training samples. In
our approach, the parameters of the model are explicitly trained such that a
small number of gradient steps with a small amount of training data from a new
task will produce good generalization performance on that task. In effect, our
method trains the model to be easy to fine-tune. We demonstrate that this
approach leads to state-of-the-art performance on two few-shot image
classification benchmarks, produces good results on few-shot regression, and
accelerates fine-tuning for policy gradient reinforcement learning with neural
network policies.
@article{finn2017modelagnostic,
abstract = {We propose an algorithm for meta-learning that is model-agnostic, in the
sense that it is compatible with any model trained with gradient descent and
applicable to a variety of different learning problems, including
classification, regression, and reinforcement learning. The goal of
meta-learning is to train a model on a variety of learning tasks, such that it
can solve new learning tasks using only a small number of training samples. In
our approach, the parameters of the model are explicitly trained such that a
small number of gradient steps with a small amount of training data from a new
task will produce good generalization performance on that task. In effect, our
method trains the model to be easy to fine-tune. We demonstrate that this
approach leads to state-of-the-art performance on two few-shot image
classification benchmarks, produces good results on few-shot regression, and
accelerates fine-tuning for policy gradient reinforcement learning with neural
network policies.},
added-at = {2019-03-08T17:27:00.000+0100},
author = {Finn, Chelsea and Abbeel, Pieter and Levine, Sergey},
biburl = {https://www.bibsonomy.org/bibtex/20ce66d088c986afe6af0f2449ea5e849/kirk86},
description = {[1703.03400] Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networks},
interhash = {52c1ed07488c96d0a0d4f19d1c4d635f},
intrahash = {0ce66d088c986afe6af0f2449ea5e849},
keywords = {meta-learning},
note = {cite arxiv:1703.03400Comment: ICML 2017. Code at https://github.com/cbfinn/maml, Videos of RL results at https://sites.google.com/view/maml, Blog post at http://bair.berkeley.edu/blog/2017/07/18/learning-to-learn/},
timestamp = {2019-03-08T17:27:00.000+0100},
title = {Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networks},
url = {http://arxiv.org/abs/1703.03400},
year = 2017
}