Summary of Accelerating Legacy Numerical Solvers by Non-intrusive Gradient-based Meta-solving, By Sohei Arisaka and Qianxiao Li
Accelerating Legacy Numerical Solvers by Non-intrusive Gradient-based Meta-solving
by Sohei Arisaka, Qianxiao Li
First submitted to arxiv on: 5 May 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: Numerical Analysis (math.NA)
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 In this paper, researchers explore ways to accelerate scientific computing by leveraging machine learning techniques for hyperparameter selection in traditional numerical methods. They propose a novel methodology that combines machine learning with legacy numerical codes without requiring modifications, using a gradient estimation technique. This approach is shown to outperform other baselines and has potential applications in accelerating established non-automatic-differentiable numerical solvers implemented in PETSc. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary Scientific computing helps scientists make new discoveries and engineers design better products. But it’s slow and expensive. Researchers are trying to speed things up using machine learning, a type of artificial intelligence. They’re experimenting with ways to use machine learning to choose the best settings for old computer codes that don’t work well with this technology. This paper shows how they can do this without changing the old codes, by estimating the gradient (a math concept). This could help people use the latest research in their own projects and make scientific computing faster and more efficient. |
Keywords
» Artificial intelligence » Hyperparameter » Machine learning
