Summary of Enhancing Semi-supervised Learning Via Representative and Diverse Sample Selection, by Qian Shao et al.
Enhancing Semi-Supervised Learning via Representative and Diverse Sample Selection
by Qian Shao, Jiangrui Kang, Qiyuan Chen, Zepeng Li, Hongxia Xu, Yiwen Cao, Jiajuan Liang, Jian Wu
First submitted to arxiv on: 18 Sep 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: Computer Vision and Pattern Recognition (cs.CV)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary A novel approach to semi-supervised learning, Representative and Diverse Sample Selection (RDSS), is proposed to tackle the under-explored issue of sample selection for labelling in deep learning tasks. By minimizing a criterion called α-Maximum Mean Discrepancy (α-MMD) using a modified Frank-Wolfe algorithm, RDSS selects a representative and diverse subset of samples from unlabeled data for annotation. This approach is shown to enhance the generalization ability of low-budget learning and consistently improves the performance of popular semi-supervised learning frameworks, outperforming state-of-the-art sample selection methods used in active learning and semi-supervised active learning. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary Semi-supervised learning helps computers learn from both labeled and unlabeled data. This approach makes it easier to train models without needing as much human labor. Researchers have been working on using unlabeled data effectively, but they haven’t paid enough attention to how to choose which samples to label. To fix this gap, scientists developed a new method called Representative and Diverse Sample Selection (RDSS). RDSS chooses the best samples from the unlabeled data to be labeled by minimizing a special criterion called α-MMD. This approach helps models learn better with limited labeling budgets. |
Keywords
» Artificial intelligence » Active learning » Attention » Deep learning » Generalization » Semi supervised
