Summary of Toward Large Kernel Models, by Amirhesam Abedsoltan et al.
Toward Large Kernel Models
by Amirhesam Abedsoltan, Mikhail Belkin, Parthe Pandit
First submitted to arxiv on: 6 Feb 2023
Categories
- Main: Machine Learning (cs.LG)
- Secondary: 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 introduces a new approach to constructing large-scale general kernel models, which decouples the model size from the training data size, allowing for scalable learning. Building upon recent studies showing kernel machines can perform similarly or better than deep neural networks on small datasets, the authors propose EigenPro 3.0, an algorithm based on projected dual preconditioned SGD. This innovation enables training on large datasets, a significant improvement over traditional kernel methods. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This paper helps us understand how to make machine learning models work well with big data. Right now, we can only use small models with small amounts of data. The authors are trying to fix this problem by creating a new type of model that can handle lots of data. They’re calling it EigenPro 3.0 and it’s based on an algorithm that helps the model learn from large datasets. |
Keywords
* Artificial intelligence * Machine learning
