Summary of Understanding the Double Descent Phenomenon in Deep Learning, by Marc Lafon and Alexandre Thomas
Understanding the Double Descent Phenomenon in Deep Learning
by Marc Lafon, Alexandre Thomas
First submitted to arxiv on: 15 Mar 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: Computer Vision and Pattern Recognition (cs.CV); Machine Learning (stat.ML)
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 The paper explores a classical strategy in machine learning that combines empirical risk minimization with capacity control to control overfitting and improve generalization performance. The authors examine how large over-parameterized models, such as neural networks, can be optimized to fit training data while still achieving great test error performance. By analyzing the interpolation point where increasing model complexity seems to lower test error, the researchers aim to better understand the relationship between model capacity and generalization gap. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This paper looks at how machine learning works when we try to make models that are really good at fitting training data. It finds that even though we might think more complex models would get worse results, they actually do well on new, unseen data. The authors want to figure out why this happens and what it means for making better predictions. |
Keywords
* Artificial intelligence * Generalization * Machine learning * Overfitting
