Gradient-EM Bayesian Meta-learning
This work addresses computational bottlenecks in meta-learning for researchers and practitioners, offering a more efficient and flexible approach, though it is incremental as it builds on existing Bayesian meta-learning frameworks.
The paper tackled the computational inefficiency and inflexibility in Bayesian meta-learning by proposing a gradient-EM variant that avoids back-propagation in meta-updates and decouples inner optimization, resulting in improved accuracy, reduced computation cost, and enhanced robustness to uncertainty in tasks like sinusoidal regression, few-shot image classification, and reinforcement learning.
Bayesian meta-learning enables robust and fast adaptation to new tasks with uncertainty assessment. The key idea behind Bayesian meta-learning is empirical Bayes inference of hierarchical model. In this work, we extend this framework to include a variety of existing methods, before proposing our variant based on gradient-EM algorithm. Our method improves computational efficiency by avoiding back-propagation computation in the meta-update step, which is exhausting for deep neural networks. Furthermore, it provides flexibility to the inner-update optimization procedure by decoupling it from meta-update. Experiments on sinusoidal regression, few-shot image classification, and policy-based reinforcement learning show that our method not only achieves better accuracy with less computation cost, but is also more robust to uncertainty.