English

Can semi-supervised learning use all the data effectively? A lower bound perspective

Machine Learning 2023-12-01 v1 Machine Learning

Abstract

Prior works have shown that semi-supervised learning algorithms can leverage unlabeled data to improve over the labeled sample complexity of supervised learning (SL) algorithms. However, existing theoretical analyses focus on regimes where the unlabeled data is sufficient to learn a good decision boundary using unsupervised learning (UL) alone. This begs the question: Can SSL algorithms simultaneously improve upon both UL and SL? To this end, we derive a tight lower bound for 2-Gaussian mixture models that explicitly depends on the labeled and the unlabeled dataset size as well as the signal-to-noise ratio of the mixture distribution. Surprisingly, our result implies that no SSL algorithm can improve upon the minimax-optimal statistical error rates of SL or UL algorithms for these distributions. Nevertheless, we show empirically on real-world data that SSL algorithms can still outperform UL and SL methods. Therefore, our work suggests that, while proving performance gains for SSL algorithms is possible, it requires careful tracking of constants.

Keywords

Cite

@article{arxiv.2311.18557,
  title  = {Can semi-supervised learning use all the data effectively? A lower bound perspective},
  author = {Alexandru Ţifrea and Gizem Yüce and Amartya Sanyal and Fanny Yang},
  journal= {arXiv preprint arXiv:2311.18557},
  year   = {2023}
}

Comments

Published in Advances in Neural Information Processing Systems 2023

R2 v1 2026-06-28T13:36:57.523Z