Meta Learning on a Sequence of Imbalanced Domains with Difficulty Awareness
This work addresses a practical challenge for real-world machine learning systems that need to adapt to non-stationary and imbalanced data streams, representing an incremental improvement over existing meta-learning methods.
The paper tackles the problem of meta-learning in evolving environments with imbalanced and shifting task distributions, proposing a kernel-based domain change detection method and difficulty-aware memory management that achieves effective continuous learning across domains.
Recognizing new objects by learning from a few labeled examples in an evolving environment is crucial to obtain excellent generalization ability for real-world machine learning systems. A typical setting across current meta learning algorithms assumes a stationary task distribution during meta training. In this paper, we explore a more practical and challenging setting where task distribution changes over time with domain shift. Particularly, we consider realistic scenarios where task distribution is highly imbalanced with domain labels unavailable in nature. We propose a kernel-based method for domain change detection and a difficulty-aware memory management mechanism that jointly considers the imbalanced domain size and domain importance to learn across domains continuously. Furthermore, we introduce an efficient adaptive task sampling method during meta training, which significantly reduces task gradient variance with theoretical guarantees. Finally, we propose a challenging benchmark with imbalanced domain sequences and varied domain difficulty. We have performed extensive evaluations on the proposed benchmark, demonstrating the effectiveness of our method. We made our code publicly available.