Corpus ID: 25316837

Meta-SGD: Learning to Learn Quickly for Few Shot Learning

@article{Li2017MetaSGDLT,
  title={Meta-SGD: Learning to Learn Quickly for Few Shot Learning},
  author={Zhenguo Li and Fengwei Zhou and F. Chen and H. Li},
  journal={ArXiv},
  year={2017},
  volume={abs/1707.09835}
}
  • Zhenguo Li, Fengwei Zhou, +1 author H. Li
  • Published 2017
  • Computer Science, Mathematics
  • ArXiv
  • Few-shot learning is challenging for learning algorithms that learn each task in isolation and from scratch. In contrast, meta-learning learns from many related tasks a meta-learner that can learn a new task more accurately and faster with fewer examples, where the choice of meta-learners is crucial. In this paper, we develop Meta-SGD, an SGD-like, easily trainable meta-learner that can initialize and adapt any differentiable learner in just one step, on both supervised learning and… CONTINUE READING
    Deep Meta-Learning: Learning to Learn in the Concept Space
    52
    Task Agnostic Meta-Learning for Few-Shot Learning
    47
    Meta-Transfer Learning for Few-Shot Learning
    135
    Few-shot Learning with Meta Metric Learners
    11
    Rethink and Redesign Meta learning
    3
    Meta-Learning in Neural Networks: A Survey
    30
    Hierarchical Meta Learning
    2

    References

    Publications referenced by this paper.
    SHOWING 1-10 OF 33 REFERENCES
    Optimization as a Model for Few-Shot Learning
    1133
    Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networks
    2155
    Meta-Learning with Temporal Convolutions
    127
    Matching Networks for One Shot Learning
    1774
    Learning to reinforcement learn
    382
    RL$^2$: Fast Reinforcement Learning via Slow Reinforcement Learning
    353
    Multitask Learning
    2010
    Siamese Neural Networks for One-Shot Image Recognition
    • Gregory R. Koch
    • Computer Science
    • 2015
    1229
    One-Shot Generalization in Deep Generative Models
    162