Generalized Reinforcement Meta Learning for Few-Shot Optimization

2020/05/04 by Raviteja Anantha, Anantha, Raviteja, Stephen Pulman +3 · 1 citation
Computer Science · #Advanced Neural Network Applications #Artificial Intelligence (cs.AI) #Domain Adaptation and Few-Shot Learning #FOS: Computer and information sciences #Machine Learning (cs.LG) #Machine Learning (stat.ML) #Multimodal Machine Learning Applications

paper · pdf · doi:10.48550/arxiv.2005.01246

openalex publication_date 2020/05/04 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28

Abstract

We present a generic and flexible Reinforcement Learning (RL) based meta-learning framework for the problem of few-shot learning. During training, it learns the best optimization algorithm to produce a learner (ranker/classifier, etc) by exploiting stable patterns in loss surfaces. Our method implicitly estimates the gradients of a scaled loss function while retaining the general properties intact for parameter updates. Besides providing improved performance on few-shot tasks, our framework could be easily extended to do network architecture search. We further propose a novel dual encoder, affinity-score based decoder topology that achieves additional improvements to performance. Experiments on an internal dataset, MQ2007, and AwA2 show our approach outperforms existing alternative approaches by 21%, 8%, and 4% respectively on accuracy and NDCG metrics. On Mini-ImageNet dataset our approach achieves comparable results with Prototypical Networks. Empirical evaluations demonstrate that our approach provides a unified and effective framework.

Citations

• Introducing LETOR 4.0 Datasets
• One-shot Learning with Memory-Augmented Neural Networks
• Learning to Optimize
• Learning to reinforcement learn
• Neural Architecture Search with Reinforcement Learning
• RL²: Fast Reinforcement Learning via Slow Reinforcement Learning
• L4: Practical loss-based stepsize adaptation for deep learning

Cited by

• Analytical Survey of Learning with Low-Resource Data: From Analysis to Investigation

Related

• Fast Few-Shot Classification by Few-Iteration Meta-Learning
• Efficient Automatic Meta Optimization Search for Few-Shot Learning
• MetAdapt: Meta-Learned Task-Adaptive Architecture for Few-Shot Classification
• Meta Navigator: Search for a Good Adaptation Policy for Few-shot Learning
• Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networks