Summary of Interval Abstractions For Robust Counterfactual Explanations, by Junqi Jiang et al.
Interval Abstractions for Robust Counterfactual Explanations
by Junqi Jiang, Francesco Leofante, Antonio Rago, Francesca Toni
First submitted to arxiv on: 21 Apr 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: Artificial Intelligence (cs.AI)
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 proposes a novel interval abstraction technique for parametric machine learning models to provide provable robustness guarantees for Counterfactual Explanations (CEs) under model changes. The approach formalizes a robustness notion called -robustness, which ensures CEs remain valid even when slight changes occur in the model parameters. The authors develop procedures using Mixed Integer Linear Programming to verify and generate -robust CEs. The paper presents an extensive empirical study demonstrating the practical applicability of this approach using neural networks and logistic regression models. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This paper helps us understand how to make sure that the reasons we get for a machine learning model’s decision are still correct even if someone tweaks the model slightly. It proposes a new way to do this called interval abstraction, which gives us guaranteed results. The authors also come up with a new idea of “robustness” to describe when our explanations stay accurate. They show how to use this concept to make more reliable explanations using special algorithms and mathematical techniques. |
Keywords
» Artificial intelligence » Logistic regression » Machine learning
