Unsupervised Data Augmentation for Consistency Training · HackerPapers discussions

Unsupervised Data Augmentation for Consistency Training — mentioned by hallqv

Fri, 04 Sep 2020 19:30:37 GMT

Agree with previous posts re: reading papers being potential rabbit hole for NLP practitioners. One paper that could be pretty useful for practical applications is this one https://arxiv.org/pdf/1904.12848.pdf It outlines strategies for data augmentation in NLP, as well as other ML task. Finding task-specific labeled data is often one of the most pertinent issues for applying ML outside of academia.

Paper and discussion history. A mention is not an endorsement.

Unsupervised Data Augmentation for Consistency Training — mentioned by amitness

Fri, 03 Apr 2020 16:26:45 GMT

Agree with you on this. For text data, there was a paper called "UDA"(https://arxiv.org/abs/1904.12848) that did some work on this direction. They augmented text by using backtranslation. Basic idea is you take text in English, translate that to some other language say French and then translate back the French text to English. Usually, you get back an English sentence that is different that the original English sentence but has the same meaning. Another approach they use to augment is to randomly replace stopword/low tf-idf(intuitively say very frequent words like a, an, the) with random…

Paper and discussion history. A mention is not an endorsement.

Unsupervised Data Augmentation for Consistency Training — HN submission

Fri, 12 Jul 2019 01:31:55 GMT

Unsupervised Data Augmentation for Consistency Training

1 point; 0 discussion replies on HN. Paper and source history. A submission is not an endorsement.

Unsupervised Data Augmentation for Consistency Training — HN submission

Tue, 30 Apr 2019 04:57:05 GMT

Unsupervised Data Augmentation

1 point; 2 discussion replies on HN. Paper and source history. A submission is not an endorsement.