Skip to content

Accompanying notebook and sources to "A Guide to Pseudolabelling: How to get a Kaggle medal with only one model" (Dec. 2020 PyData Boston-Cambridge Keynote)

Notifications You must be signed in to change notification settings

stanleyjzheng/PyData-Pseudolabelling-Keynote

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

8 Commits
 
 
 
 
 
 

Repository files navigation

A Guide to Pseudolabelling: How to get a Kaggle medal with only one model

PyData Boston-Cambridge Dec. 2020

Talk now on YouTube: https://youtu.be/c8uWUOSGYUI

Notebook on Kaggle: https://www.kaggle.com/stanleyjzheng/exploring-pseudolabelling-schemes-pydata

Me

References

[1] Babakhin, Y., Sanakoyeu, A., & Kitamura, H. (2019). Semi-supervised segmentation of salt bodies in seismic images using an ensemble of convolutional neural networks. ArXiv:1904.04445. http://arxiv.org/abs/1904.04445

[2] Internet.org and Facebook (2013). A Focus on Efficiency. [White paper]. Archived on webarchive.org

[3] Li, Z., Ko, B., & Choi, H.-J. (2019). Naive semi-supervised deep learning using pseudo-label. Peer-to-Peer Networking and Applications, 12(5), 1358–1368. https://doi.org/10.1007/s12083-018-0702-9

  • Paper describing pretrain method for pseudolabelling as well as results from LSTM, CIFAR, MNIST.

[4] Xie, Q., Luong, M.-T., Hovy, E., & Le, Q. V. (2020). Self-training with Noisy Student improves ImageNet classification. ArXiv:1911.04252 [Cs, Stat]. http://arxiv.org/abs/1911.04252

[5] Global Wheat competition: https://www.kaggle.com/c/global-wheat-detection

[6] The following are solutions from the mentioned competitions: OpenVaccine 1st, OpenVaccine 2nd, OpenVaccine 3rd, Tweet sentiment extraction 1st, TReNDS Neuroimaging 1st, Global Wheat 1st, LISH-MOA 2nd public, LISH-MOA 5th, TGS Salt Identification 1st

[7] My notebook on pseudolabelling MNIST https://www.kaggle.com/stanleyjzheng/exploring-pseudolabelling-schemes-pydata

Further reading

About

Accompanying notebook and sources to "A Guide to Pseudolabelling: How to get a Kaggle medal with only one model" (Dec. 2020 PyData Boston-Cambridge Keynote)

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published