Summary of Debiasing Mini-batch Quadratics For Applications in Deep Learning, by Lukas Tatzel et al.
Debiasing Mini-Batch Quadratics for Applications in Deep Learning
by Lukas Tatzel, Bálint Mucsányi, Osane Hackel, Philipp Hennig
First submitted to arxiv on: 18 Oct 2024
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 This paper explores the impact of mini-batching on quadratic approximations used in machine learning. Specifically, it examines how computing stochastic quadratic approximations on mini-batches can introduce a systematic error that affects applications like second-order optimization and uncertainty quantification via the Laplace approximation. The authors provide a theoretical explanation for this bias and develop debiasing strategies to mitigate its effects. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This paper is about how machine learning works when we don’t have enough computing power. Currently, we use small groups of data (mini-batches) instead of the whole dataset. This can cause problems with certain types of calculations used in deep learning. The authors show that these problems can lead to incorrect results and develop ways to fix them. |
Keywords
* Artificial intelligence * Deep learning * Machine learning * Optimization
