No abstract

No abstract

No abstract

Semi-supervised learning (SSL) algorithms have had great success in recent years in limited labeled data regimes. However, the current state-of-the-art SSL algorithms are computationally expensive and entail significant compute time and energy requirements. This can prove to be a huge limitation for many smaller companies and academic groups. Our main insight is that training on a subset of unlabeled data instead of entire unlabeled data enables the current SSL algorithms to converge faster, thereby reducing the computational costs significantly. In this work, we propose RETRIEVE 1 , a coreset selection framework for efficient and robust semi-supervised learning. RETRIEVE selects the coreset by solving a mixed discrete-continuous bi-level optimization problem such that the selected coreset minimizes the labeled set loss. We use a one-step gradient approximation and show that the discrete optimization problem is approximately submodular, thereby enabling simple greedy algorithms to obtain the coreset. We empirically demonstrate on several real-world datasets that existing SSL algorithms like VAT, Mean-Teacher, FixMatch, when used with RETRIEVE, achieve a) faster training times, b) better performance when unlabeled data consists of Out-of-Distribution (OOD) data and imbalance. More specifically, we show that with minimal accuracy degradation, RETRIEVE achieves a speedup of around 3× in the traditional SSL setting and achieves a speedup of 5× compared to state-of-the-art (SOTA) robust SSL algorithms in the case of imbalance and OOD data.However, the current SOTA SSL algorithms are compute-intensive with very large training times. For example, from our personal experience, training a WideResNet model [52] on a CIFAR10 [24] 1 coResets for EfficienT and Robust semI-supErVised lEarning Preprint. Under review.

Model-Agnostic Meta-Learning (MAML) is a popular gradient-based meta-learning framework that tries to find an optimal initialization to minimize the expected loss across all tasks during metatraining. However, it inherently assumes that the contribution of each instance/task to the meta-learner is equal. Therefore, it fails to address the problem of domain differences between base and novel classes in few-shot learning. In this work, we propose a novel and robust meta-learning algorithm, called RW-MAML, which learns to assign weights to training instances or tasks. We consider these weights to be hyper-parameters. Hence, we iteratively optimize the weights using a small set of validation tasks and an online approximation in a bi-bi-level optimization framework, in contrast to the standard bi-level optimization in MAML. Therefore, we investigate a practical evaluation setting to demonstrate the scalability of our RW-MAML in two scenarios: (1) out-of-distribution tasks and (2) noisy labels in the meta-training stage. Extensive experiments on synthetic and real-world datasets demonstrate that our framework efficiently mitigates the effects of "unwanted" instances, showing that our proposed technique significantly outperforms state-of-the-art robust meta-learning methods.