Abstract | ||
---|---|---|
Semi-supervised learning has proven to be a powerful paradigm for leveraging unlabeled data to mitigate the reliance on large labeled datasets.In this work, we unify the current dominant approaches for semi-supervised learning to produce a new algorithm, MixMatch, thatguesses low-entropy labels for data-augmented unlabeled examples and mixes labeled and unlabeled data using MixUp.MixMatch obtains state-of-the-art results by a large margin across many datasets and labeled data amounts. For example,on CIFAR-10 with 250 labels, we reduce error rate by a factor of 4 (from 38% to 11%) and by a factor of 2 on STL-10.We also demonstrate how MixMatch can help achieve a dramatically better accuracy-privacy trade-off for differential privacy.Finally, we perform an ablation study to tease apart which components of MixMatch are most important for its success.Code is attached. |
Year | Venue | DocType |
---|---|---|
2019 | NeurIPS | Conference |
Citations | PageRank | References |
0 | 0.34 | 0 |
Authors | ||
6 |
Name | Order | Citations | PageRank |
---|---|---|---|
David Berthelot | 1 | 209 | 10.40 |
Nicholas Carlini | 2 | 1599 | 63.23 |
Ian J. Goodfellow | 3 | 5224 | 268.13 |
Nicolas Papernot | 4 | 1932 | 87.62 |
Avital Oliver | 5 | 53 | 4.69 |
Colin Raffel | 6 | 190 | 21.50 |