Skip to content
master
Switch branches/tags
Code

Latest commit

 

Git stats

Files

Permalink
Failed to load latest commit information.
Type
Name
Latest commit message
Commit time
 
 
 
 

NS3L

This repository contains a python implementation of Negative Sampling in Semi-Supervised Learning. The experimental section in the paper can be reproduced using code in this repository. The original paper is due to appear at ICML 2020, and the arxiv version is viewable at https://arxiv.org/pdf/1911.05166.pdf.

Reference

Negative Sampling in Semi-Supervised Learning (NS3L) was proposed in the ICML 2020 paper "Negative Sampling in Semi-Supervised Learning". We recommend reading the paper for the motivation and empirical results of this method. NS3L is designed to boost the performance of existing SSL algorithms simply by adding the loss and is fairly robust to a reasonably large range of hyperparameters.

If this algorithm is useful, please cite the paper as follows:

@article{chen2020negative,
  title={Negative Sampling in Semi-Supervised Learning},
  author={John Chen, Vatsal Shah, and Anastasios Kyrillidis},
  journal={arXiv preprint arXiv:1911.05166},
  year={2020}
}

About

Implementation for Negative Sampling in Semi-Supervised Learning

Resources

Releases

No releases published

Packages

No packages published

Languages