English

More Than Meets The Eye: Semi-supervised Learning Under Non-IID Data

Machine Learning 2021-04-27 v1 Computer Vision and Pattern Recognition Machine Learning

Abstract

A common heuristic in semi-supervised deep learning (SSDL) is to select unlabelled data based on a notion of semantic similarity to the labelled data. For example, labelled images of numbers should be paired with unlabelled images of numbers instead of, say, unlabelled images of cars. We refer to this practice as semantic data set matching. In this work, we demonstrate the limits of semantic data set matching. We show that it can sometimes even degrade the performance for a state of the art SSDL algorithm. We present and make available a comprehensive simulation sandbox, called non-IID-SSDL, for stress testing an SSDL algorithm under different degrees of distribution mismatch between the labelled and unlabelled data sets. In addition, we demonstrate that simple density based dissimilarity measures in the feature space of a generic classifier offer a promising and more reliable quantitative matching criterion to select unlabelled data before SSDL training.

Keywords

Cite

@article{arxiv.2104.10223,
  title  = {More Than Meets The Eye: Semi-supervised Learning Under Non-IID Data},
  author = {Saul Calderon-Ramirez and Luis Oala},
  journal= {arXiv preprint arXiv:2104.10223},
  year   = {2021}
}

Comments

Presented as a RobustML workshop paper at ICLR 2021. Both authors contributed equally. This article extends arXiv:2006.07767

R2 v1 2026-06-24T01:22:58.164Z