arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 1904.09545cs.CL
收藏

足够好的成分数据增强

Good-Enough Compositional Data Augmentation

MIT CSAIL(麻省理工学院计算机科学与人工智能实验室)

Jacob Andreas

摘要:

我们提出了一个简单的数据增强协议,旨在提供一个组成的归纳偏见的条件和无条件的序列模型。在该协议下,合成的训练样本是通过采用真实的训练样本并用出现在至少一个类似环境中的其他片段替换(可能是不连续的)片段来构建的。该协议是模型不可知的,适用于各种任务。应用于神经序列到序列模型,它将SCAN数据集的诊断任务的出错率降低了87%,将语义解析任务的错误率降低了16%。应用于n-gram语言模型,它在几种语言的小型语料库上减少了大约1%的困惑。

英文摘要:

We propose a simple data augmentation protocol aimed at providing a compositional inductive bias in conditional and unconditional sequence models. Under this protocol, synthetic training examples are constructed by taking real training examples and replacing (possibly discontinuous) fragments with other fragments that appear in at least one similar environment. The protocol is model-agnostic and useful for a variety of tasks. Applied to neural sequence-to-sequence models, it reduces error rate by as much as 87% on diagnostic tasks from the SCAN dataset and 16% on a semantic parsing task. Applied to n-gram language models, it reduces perplexity by roughly 1% on small corpora in several languages.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑