Summary of Conformal-in-the-loop For Learning with Imbalanced Noisy Data, by John Brandon Graham-knight et al.
Conformal-in-the-Loop for Learning with Imbalanced Noisy Data
by John Brandon Graham-Knight, Jamil Fayyad, Nourhan Bayasi, Patricia Lasserre, Homayoun Najjaran
First submitted to arxiv on: 4 Nov 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 This research paper proposes Conformal-in-the-Loop (CitL), a novel training framework that addresses class imbalance and label noise simultaneously. The existing approaches typically address one or the other issue, leading to suboptimal results when both issues coexist. CitL evaluates sample uncertainty to adjust weights and prune unreliable examples, enhancing model resilience and accuracy with minimal computational cost. The authors conducted extensive experiments, showing how CitL effectively emphasizes impactful data in noisy, imbalanced datasets. Their results demonstrate that CitL consistently boosts model performance, achieving up to a 6.1% increase in classification accuracy and a 5.0 mIoU improvement in segmentation. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This paper proposes a new way to train machine learning models when the training data is messy and biased. Right now, most research assumes that the data is clean and balanced, which isn’t true for many real-world problems. The authors introduce Conformal-in-the-Loop (CitL), a method that can handle both class imbalance and label noise at the same time. They tested CitL on many different datasets and showed that it works well, even when the data is very noisy or imbalanced. |
Keywords
* Artificial intelligence * Classification * Machine learning
