Robust MAML: Prioritization task buffer with adaptive learning process for model-agnostic meta-learning

TL;DR

Robust MAML introduces an adaptive learning scheme and prioritization task buffer to enhance scalability, reduce hyper-parameter sensitivity, and improve robustness to distribution mismatch in meta-learning tasks.

Contribution

The paper proposes Robust MAML, which automatically optimizes learning rates and gradually aligns training and testing task distributions for better meta-learning performance.

Findings

Significant performance improvements in meta reinforcement learning environments.

Reduced sensitivity to hyper-parameter tuning.

Enhanced robustness to distribution mismatch.

Abstract

Model agnostic meta-learning (MAML) is a popular state-of-the-art meta-learning algorithm that provides good weight initialization of a model given a variety of learning tasks. The model initialized by provided weight can be fine-tuned to an unseen task despite only using a small amount of samples and within a few adaptation steps. MAML is simple and versatile but requires costly learning rate tuning and careful design of the task distribution which affects its scalability and generalization. This paper proposes a more robust MAML based on an adaptive learning scheme and a prioritization task buffer(PTB) referred to as Robust MAML (RMAML) for improving scalability of training process and alleviating the problem of distribution mismatch. RMAML uses gradient-based hyper-parameter optimization to automatically find the optimal learning rate and uses the PTB to gradually adjust train-ing task distribution toward testing task distribution over the course of training. Experimental results on meta reinforcement learning environments demonstrate a substantial performance gain as well as being less sensitive to hyper-parameter choice and robust to distribution mismatch.