Few-Shot Learning is an example of meta-learning, where a learner is trained on several related tasks, during the meta-training phase, so that it can generalize well to unseen (but related) tasks with just few examples, during the meta-testing phase. An effective approach to the Few-Shot Learning problem is to learn a common representation for various tasks and train task specific classifiers on top of this representation.
|TREND||DATASET||BEST METHOD||PAPER TITLE||PAPER||CODE||COMPARE|
By contrast, humans can generally perform a new language task from only a few examples or from simple instructions - something which current NLP systems still largely struggle to do.
Ranked #1 on Sentence Completion on HellaSwag
COMMON SENSE REASONING COREFERENCE RESOLUTION DOMAIN ADAPTATION FEW-SHOT LEARNING LANGUAGE MODELLING MULTI-TASK LEARNING NATURAL LANGUAGE INFERENCE QUESTION ANSWERING SENTENCE COMPLETION UNSUPERVISED MACHINE TRANSLATION WORD SENSE DISAMBIGUATION
We propose an algorithm for meta-learning that is model-agnostic, in the sense that it is compatible with any model trained with gradient descent and applicable to a variety of different learning problems, including classification, regression, and reinforcement learning.
In this paper, we present FewJoint, a novel Few-Shot Learning benchmark for NLP.
We conclude with a discussion of the rapid learning vs feature reuse question for meta-learning algorithms more broadly.
This paper considers meta-learning problems, where there is a distribution of tasks, and we would like to obtain an agent that performs well (i. e., learns quickly) when presented with a previously unseen task sampled from this distribution.
In contrast, meta-learning learns from many related tasks a meta-learner that can learn a new task more accurately and faster with fewer examples, where the choice of meta-learners is crucial.
We propose prototypical networks for the problem of few-shot classification, where a classifier must generalize to new classes not seen in the training set, given only a small number of examples of each new class.
We show that the proposed approach outperforms GBML with standard DNNs on the Omniglot dataset when the number of past tasks for meta-training is small.
We conduct detailed analysis of the main components that lead to high transfer performance.
Ranked #1 on Image Classification on CIFAR-100 (using extra training data)