Article

Free access

Model-agnostic meta-learning for fast adaptation of deep networks

Authors:

Sergey LevineAuthors Info & Claims

ICML'17: Proceedings of the 34th International Conference on Machine Learning - Volume 70

Pages 1126 - 1135

Published: 06 August 2017 Publication History

PDF eReader Publisher Site

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.

References

[1]

Abadi, Martín, Agarwal, Ashish, Barham, Paul, Brevdo, Eugene, Chen, Zhifeng, Citro, Craig, Corrado, Greg S, Davis, Andy, Dean, Jeffrey, Devin, Matthieu, et al. Tensorflow: Large-scale machine learning on heterogeneous distributed systems. arXiv preprint arXiv:1603.04467, 2016.

[2]

Andrychowicz, Marcin, Denil, Misha, Gomez, Sergio, Hoffman, Matthew W, Pfau, David, Schaul, Tom, and de Freitas, Nando. Learning to learn by gradient descent by gradient descent. In Neural Information Processing Systems (NIPS), 2016.

Digital Library

[3]

Bengio, Samy, Bengio, Yoshua, Cloutier, Jocelyn, and Gecsei, Jan. On the optimization of a synaptic learning rule. In Optimality in Artificial and Biological Neural Networks, pp. 6-8, 1992.

[4]

Bengio, Yoshua, Bengio, Samy, and Cloutier, Jocelyn. Learning a synaptic learning rule. Université de Montréal, Département d'informatique et de recherche opérationnelle, 1990.

[5]

Donahue, Jeff, Jia, Yangqing, Vinyals, Oriol, Hoffman, Judy, Zhang, Ning, Tzeng, Eric, and Darrell, Trevor. Decaf: A deep convolutional activation feature for generic visual recognition. In International Conference on Machine Learning (ICML), 2014.

Digital Library

[6]

Duan, Yan, Chen, Xi, Houthooft, Rein, Schulman, John, and Abbeel, Pieter. Benchmarking deep reinforcement learning for continuous control. In International Conference on Machine Learning (ICML), 2016a.

Digital Library

[7]

Duan, Yan, Schulman, John, Chen, Xi, Bartlett, Peter L, Sutskever, Ilya, and Abbeel, Pieter. Rl2: Fast reinforcement learning via slow reinforcement learning. arXiv preprint arXiv:1611.02779, 2016b.

[8]

Edwards, Harrison and Storkey, Amos. Towards a neural statistician. International Conference on Learning Representations (ICLR), 2017.

[9]

Goodfellow, Ian J, Shlens, Jonathon, and Szegedy, Christian. Explaining and harnessing adversarial examples. International Conference on Learning Representations (ICLR), 2015.

[10]

Ha, David, Dai, Andrew, and Le, Quoc V. Hypernetworks. International Conference on Learning Representations (ICLR), 2017.

[11]

Hochreiter, Sepp, Younger, A Steven, and Conwell, Peter R. Learning to learn using gradient descent. In International Conference on Artificial Neural Networks. Springer, 2001.

Digital Library

[12]

Husken, Michael and Goerick, Christian. Fast learning for problem classes using knowledge based network initialization. In Neural Networks, 2000. IJCNN 2000, Proceedings of the IEEE-INNS-ENNS International Joint Conference on, volume 6, pp. 619-624. IEEE, 2000.

Digital Library

[13]

Ioffe, Sergey and Szegedy, Christian. Batch normalization: Accelerating deep network training by reducing internal covariate shift. International Conference on Machine Learning (ICML), 2015.

Digital Library

[14]

Kaiser, Lukasz, Nachum, Ofir, Roy, Aurko, and Bengio, Samy. Learning to remember rare events. International Conference on Learning Representations (ICLR), 2017.

[15]

Kingma, Diederik and Ba, Jimmy. Adam: A method for stochastic optimization. International Conference on Learning Representations (ICLR), 2015.

[16]

Kirkpatrick, James, Pascanu, Razvan, Rabinowitz, Neil, Veness, Joel, Desjardins, Guillaume, Rusu, Andrei A, Milan, Kieran, Quan, John, Ramalho, Tiago, Grabska-Barwinska, Agnieszka, et al. Overcoming catastrophic forgetting in neural networks. arXiv preprint arXiv:1612.00796, 2016.

[17]

Koch, Gregory. Siamese neural networks for one-shot image recognition. ICML Deep Learning Workshop, 2015.

[18]

Krähenbuhl, Philipp, Doersch, Carl, Donahue, Jeff, and Darrell, Trevor. Data-dependent initializations of convolutional neural networks. International Conference on Learning Representations (ICLR), 2016.

[19]

Lake, Brenden M, Salakhutdinov, Ruslan, Gross, Jason, and Tenenbaum, Joshua B. One shot learning of simple visual concepts. In Conference of the Cognitive Science Society (CogSci), 2011.

[20]

Li, Ke and Malik, Jitendra. Learning to optimize. International Conference on Learning Representations (ICLR), 2017.

[21]

Maclaurin, Dougal, Duvenaud, David, and Adams, Ryan. Gradient-based hyperparameter optimization through reversible learning. In International Conference on Machine Learning (ICML), 2015.

Digital Library

[22]

Munkhdalai, Tsendsuren and Yu, Hong. Meta networks. International Conferecence on Machine Learning (ICML), 2017.

[23]

Naik, Devang K and Mammone, RJ. Meta-neural networks that learn by learning. In International Joint Conference on Neural Netowrks (IJCNN), 1992.

[24]

Parisotto, Emilio, Ba, Jimmy Lei, and Salakhutdinov, Ruslan. Actor-mimic: Deep multitask and transfer reinforcement learning. International Conference on Learning Representations (ICLR), 2016.

[25]

Ravi, Sachin and Larochelle, Hugo. Optimization as a model for few-shot learning. In International Conference on Learning Representations (ICLR), 2017.

[26]

Rei, Marek. Online representation learning in recurrent neural language models. arXiv preprint arXiv:1508.03854, 2015.

[27]

Rezende, Danilo Jimenez, Mohamed, Shakir, Danihelka, Ivo, Gregor, Karol, and Wierstra, Daan. One-shot generalization in deep generative models. International Conference on Machine Learning (ICML), 2016.

Digital Library

[28]

Salimans, Tim and Kingma, Diederik P. Weight normalization: A simple reparameterization to accelerate training of deep neural networks. In Neural Information Processing Systems (NIPS), 2016.

Digital Library

[29]

Santoro, Adam, Bartunov, Sergey, Botvinick, Matthew, Wierstra, Daan, and Lillicrap, Timothy. Meta-learning with memory-augmented neural networks. In International Conference on Machine Learning (ICML), 2016.

Digital Library

[30]

Saxe, Andrew, McClelland, James, and Ganguli, Surya. Exact solutions to the nonlinear dynamics of learning in deep linear neural networks. International Conference on Learning Representations (ICLR), 2014.

[31]

Schmidhuber, Jurgen. Evolutionary principles in self-referential learning. On learning how to learn: The meta-meta-... hook.) Diploma thesis, Institut f. Informatik, Tech. Univ. Munich, 1987.

[32]

Schmidhuber, Jürgen. Learning to control fast-weight memories: An alternative to dynamic recurrent networks. Neural Computation, 1992.

Digital Library

[33]

Schulman, John, Levine, Sergey, Abbeel, Pieter, Jordan, Michael I, and Moritz, Philipp. Trust region policy optimization. In International Conference on Machine Learning (ICML), 2015.

Digital Library

[34]

Shyam, Pranav, Gupta, Shubham, and Dukkipati, Ambedkar. Attentive recurrent comparators. International Conferecence on Machine Learning (ICML), 2017.

[35]

Snell, Jake, Swersky, Kevin, and Zemel, Richard S. Prototypical networks for few-shot learning. arXiv preprint arXiv:1703.05175, 2017.

[36]

Thrun, Sebastian and Pratt, Lorien. Learning to learn. Springer Science & Business Media, 1998.

Digital Library

[37]

Todorov, Emanuel, Erez, Tom, and Tassa, Yuval. Mujoco: A physics engine for model-based control. In International Conference on Intelligent Robots and Systems (IROS), 2012.

[38]

Vinyals, Oriol, Blundell, Charles, Lillicrap, Tim, Wierstra, Daan, et al. Matching networks for one shot learning. In Neural Information Processing Systems (NIPS), 2016.

Digital Library

[39]

Wang, Jane X, Kurth-Nelson, Zeb, Tirumala, Dhruva, Soyer, Hubert, Leibo, Joel Z, Munos, Remi, Blundell, Charles, Kumaran, Dharshan, and Botvinick, Matt. Learning to reinforcement learn. arXiv preprint arXiv:1611.05763, 2016.

[40]

Williams, Ronald J. Simple statistical gradient-following algorithms for connectionist reinforcement learning. Machine learning, 8(3-4):229-256, 1992.

Digital Library

Cited By

Lu JRuan JJiang HLi ZMao HZhao RDastani MSichman JAlechina NDignum V(2024)DuaLight: Enhancing Traffic Signal Control by Leveraging Scenario-Specific and Scenario-Shared KnowledgeProceedings of the 23rd International Conference on Autonomous Agents and Multiagent Systems10.5555/3635637.3662986(1283-1291)Online publication date: 6-May-2024
https://dl.acm.org/doi/10.5555/3635637.3662986
Li XXiao JChen GFeng YYang YLiu AChen L(2024)Decomposed Prototype Learning for Few-Shot Scene Graph GenerationACM Transactions on Multimedia Computing, Communications, and Applications10.1145/370087721:1(1-24)Online publication date: 21-Oct-2024
https://dl.acm.org/doi/10.1145/3700877
Xu SLuo YShi WXu SLuo YDauwels JKhong AWang ZChen QCai CShi WChua T(2024)Geo-LLaVA: A Large Multi-Modal Model for Solving Geometry Math Problems with Meta In-Context LearningProceedings of the 2nd Workshop on Large Generative Models Meet Multimodal Applications10.1145/3688866.3689124(11-15)Online publication date: 28-Oct-2024
https://dl.acm.org/doi/10.1145/3688866.3689124
Show More Cited By

Model-agnostic meta-learning for fast adaptation of deep networks
1. Computing methodologies
  1. Machine learning
    1. Machine learning approaches

Recommendations

Probabilistic model-agnostic meta-learning
NIPS'18: Proceedings of the 32nd International Conference on Neural Information Processing Systems

Meta-learning for few-shot learning entails acquiring a prior over previous tasks and experiences, such that new tasks be learned from small amounts of data. However, a critical challenge in few-shot learning is task ambiguity: even when a powerful ...
Bayesian model-agnostic meta-learning
NIPS'18: Proceedings of the 32nd International Conference on Neural Information Processing Systems

Due to the inherent model uncertainty, learning to infer Bayesian posterior from a few-shot dataset is an important step towards robust meta-learning. In this paper, we propose a novel Bayesian model-agnostic meta-learning method. The proposed method ...
Kronecker-factored Approximate Curvature with adaptive learning rate for optimizing model-agnostic meta-learning
Abstract
Model-agnostic meta-learning (MAML) highlights the ability to quickly adapt to new tasks with only a small amount of labeled training data among many few-shot learning algorithms. However, the computational complexity is high, because the MAML ...

Comments

Information & Contributors

Information

Published In

cover image Guide Proceedings

ICML'17: Proceedings of the 34th International Conference on Machine Learning - Volume 70

August 2017

4208 pages

Publisher

JMLR.org

Publication History

Published: 06 August 2017

Qualifiers

Article

Contributors

Other Metrics

View Article Metrics

Bibliometrics & Citations

Bibliometrics

Article Metrics

558
Total Citations
View Citations
7,572
Total Downloads

Downloads (Last 12 months)871
Downloads (Last 6 weeks)143

Reflects downloads up to 24 Dec 2024

Other Metrics

View Author Metrics

Citations

Cited By

Lu JRuan JJiang HLi ZMao HZhao RDastani MSichman JAlechina NDignum V(2024)DuaLight: Enhancing Traffic Signal Control by Leveraging Scenario-Specific and Scenario-Shared KnowledgeProceedings of the 23rd International Conference on Autonomous Agents and Multiagent Systems10.5555/3635637.3662986(1283-1291)Online publication date: 6-May-2024
https://dl.acm.org/doi/10.5555/3635637.3662986
Li XXiao JChen GFeng YYang YLiu AChen L(2024)Decomposed Prototype Learning for Few-Shot Scene Graph GenerationACM Transactions on Multimedia Computing, Communications, and Applications10.1145/370087721:1(1-24)Online publication date: 21-Oct-2024
https://dl.acm.org/doi/10.1145/3700877
Xu SLuo YShi WXu SLuo YDauwels JKhong AWang ZChen QCai CShi WChua T(2024)Geo-LLaVA: A Large Multi-Modal Model for Solving Geometry Math Problems with Meta In-Context LearningProceedings of the 2nd Workshop on Large Generative Models Meet Multimodal Applications10.1145/3688866.3689124(11-15)Online publication date: 28-Oct-2024
https://dl.acm.org/doi/10.1145/3688866.3689124
Buehler MLi GWood EHelminger LChen XShah TWang DGarbin SOrts-Escolano SHilliges OLagun DRiviere JGotardo PBeeler TMeka ASarkar K(2024)Cafca: High-quality Novel View Synthesis of Expressive Faces from Casual Few-shot CapturesSIGGRAPH Asia 2024 Conference Papers10.1145/3680528.3687580(1-12)Online publication date: 3-Dec-2024
https://dl.acm.org/doi/10.1145/3680528.3687580
Arbaoui MBrahmia MRahmoun AZghal M(2024)Federated Learning Survey: A Multi-Level Taxonomy of Aggregation Techniques, Experimental Insights, and Future FrontiersACM Transactions on Intelligent Systems and Technology10.1145/367818215:6(1-69)Online publication date: 20-Nov-2024
https://dl.acm.org/doi/10.1145/3678182
Pei JYan GDe Rijke MRen P(2024)Mixture-of-Languages Routing for Multilingual DialoguesACM Transactions on Information Systems10.1145/367695642:6(1-33)Online publication date: 5-Aug-2024
https://dl.acm.org/doi/10.1145/3676956
Wang SCai FZhang M(2024)Multilingual Event Causality Identification via Meta-learning with KnowledgeProceedings of the 2024 International Conference on Generative Artificial Intelligence and Information Security10.1145/3665348.3665353(18-23)Online publication date: 10-May-2024
https://dl.acm.org/doi/10.1145/3665348.3665353
Ji SLi XSun WDong HTaalas AZhang YWu HPitkänen EMarttinen P(2024)A Unified Review of Deep Learning for Automated Medical CodingACM Computing Surveys10.1145/366461556:12(1-41)Online publication date: 17-May-2024
https://dl.acm.org/doi/10.1145/3664615
Jin XLi NKong WTang JYang B(2024)Unbiased Semantic Representation Learning Based on Causal Disentanglement for Domain GeneralizationACM Transactions on Multimedia Computing, Communications, and Applications10.1145/365995320:8(1-20)Online publication date: 24-Apr-2024
https://dl.acm.org/doi/10.1145/3659953
Gharoun HMomenifar FChen FGandomi A(2024)Meta-learning Approaches for Few-Shot Learning: A Survey of Recent AdvancesACM Computing Surveys10.1145/365994356:12(1-41)Online publication date: 25-Jul-2024
https://dl.acm.org/doi/10.1145/3659943
Show More Cited By

View Options

View options

PDF

View or Download as a PDF file.

eReader

View online with eReader.

Login options

Check if you have access through your login credentials or your institution to get full access on this article.

Full Access

Get this Publication

Media

Figures

Other

Tables

View Table of Contents