Summary of Sparse and Faithful Explanations Without Sparse Models, by Yiyang Sun et al.
Sparse and Faithful Explanations Without Sparse Models
by Yiyang Sun, Zhi Chen, Vittorio Orlandi, Tong Wang, Cynthia Rudin
First submitted to arxiv on: 15 Feb 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 introduces the Sparse Explanation Value (SEV), a novel approach to measuring sparsity in machine learning models. Contrary to traditional notions of sparsity, SEV focuses on decision-making processes rather than overall model architecture. By analyzing movements over a hypercube, the authors demonstrate that many ML models, even non-sparse ones, exhibit low decision sparsity when measured by SEV. To achieve this, they propose algorithms that reduce SEV without compromising accuracy, providing faithful and sparse explanations. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This paper is about a new way to understand how machine learning models make decisions. Even if the model itself isn’t very simple, it’s possible for the reasons behind its choices to be explained using just a few key features. For example, a loan might be rejected because someone has no credit history, which overrides any other factors that might suggest they’re good borrowers. The authors introduce the concept of “decision sparsity” and show that many models, even complex ones, make decisions that can be easily understood by focusing on just a few key factors. |
Keywords
* Artificial intelligence * Machine learning
