Learning Algorithms for Active Learning

By Philip Bachman, Alessandro Sordoni, Adam Trischler. 2017

164 citations recorded by the source corpus.

We introduce a model that learns active learning algorithms via metalearning. For a distribution of related tasks, our model jointly learns: a data representation, an item selection heuristic, and a prediction function. Our model uses the item selection heuristic to construct a labeled support set for training the prediction function. Using the Omniglot and MovieLens datasets, we test our model in synthetic and practical settings.

Subfields: Artificial Intelligence, Meta & Continual Learning

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Meta & Continual Learning

Citations in the corpus

Meta-Learning for Semi-Supervised Few-Shot Classification (2018, background): “Lastly, closely related to our work with regard to extending the few-shot learning setting, Bachman et al. (2017) employ Matching Networks in an active learning framework where the model has a choice of which unlabeled item to add to the support set over a certain number of time steps before…”
Learning Rapid-Temporal Adaptations (2017, background): “There is growing interest and progress in building ﬂexible, adaptive models, particularly within the framework of metalearning , or learning to learn [24, 1, 42, 2].”
Meta-Learning in Neural Networks: A Survey (2020, background): “query policy, (iv) performing outer-level optimization to train the optimal annotation query policy [187], [188], [189].”
Generalizing from a Few Examples (2019, background): “(1) Matching Nets [138] and its variants [4, 8, 24]: Matching Nets [138] meta-learns different embedding functions (f and д) for the training sample xi and test sample xtest.”
Revisiting Meta-Learning as Supervised Learning (2020, background): “…learned with the best N ′ queried examples from Dtr, and lmeta measures the performance gap on classification. g is a learning process constrained to query labels for N ′ examples by investigating x and the learning progress (Ravi and Larochelle, 2018; Sharma et al., 2018; Bachman et al., 2017).”
A Survey on Active Learning and Human-in-the-Loop Deep Learning for Medical Image Analysis (2019, background): “Bachman et al. (2017) take this idea a step further and propose”
Rapid Adaptation with Conditionally Shifted Neurons (2017, background): “There is growing interest and progress in building flexible, adaptive models, particularly within the framework of metalearning (learning to learn) (Mitchell et al., 1993; Andrychowicz et al., 2016; Vinyals et al., 2016; Bachman et al., 2017).”
A Meta-Learning Perspective on Cold-Start Recommendations for Items (2017, methodology): “recently, [4] proposes to learn active learning algorithms via a technique based on MatchingNets.”
What Do I Annotate Next? An Empirical Study of Active Learning for Action Localization (2018, background): “Inspired by very recent ideas in active learning [1,25,40,70,74], our proposed active learning framework learns (or bootstrap) a function that selects samples for annotation based on knowledge extracted from a previous dataset.”
How to Train Your MAML to Excel in Few-Shot Classification (2021, background): “Recent years have witnessed an abundance of new approaches on meta-learning [29, 30, 48, 61, 63, 64, 68, 77, 79, 80, 94, 98], developed in various areas including few-shot learning [70, 85, 89, 105, 107, 110, 111, 115], optimization [3, 9, 53, 108], reinforcement and imitation learning [23, 24, 31, ”
Learning to Actively Learn Neural Machine Translation (2018, methodology): “The AL strategy can be learned by attempting to actively learn on tasks sampled from a distribution over the tasks (Bachman et al., 2017).”
Learning to Teach in Cooperative Multiagent Reinforcement Learning (2018, background): “Several works meta-learn active learning policies for supervised learning (Bachman et al., 2017; Fang et al., 2017; Pang et al., 2018; Fan et al., 2018).”