Summary of Introducing User Feedback-based Counterfactual Explanations (ufce), by Muhammad Suffian et al.
Introducing User Feedback-based Counterfactual Explanations (UFCE)
by Muhammad Suffian, Jose M. Alonso-Moral, Alessandro Bogliolo
First submitted to arxiv on: 26 Feb 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: Artificial Intelligence (cs.AI); Human-Computer Interaction (cs.HC)
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 A novel methodology for generating counterfactual explanations (CEs) in eXplainable Artificial Intelligence (XAI) is introduced, addressing limitations of current CE algorithms. The proposed approach, called User Feedback-based Counterfactual Explanation (UFCE), incorporates user constraints to determine the smallest modifications in a subset of actionable features while considering feature dependence and evaluating suggested changes using benchmark evaluation metrics. This methodology outperforms two well-known CE methods in terms of proximity, sparsity, and feasibility across five datasets. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary Machine learning models are used in many real-world applications, but it’s hard to understand why they make certain decisions. Counterfactual explanations can help by providing actionable information on how to achieve a desired outcome with small changes. However, current methods often ignore the most important factors that contribute to an outcome and suggest impractical changes. This study introduces UFCE, which allows users to provide constraints to find the smallest possible changes in specific features while considering how those features are connected. The method is tested on five datasets and outperforms other CE methods. |
Keywords
* Artificial intelligence * Machine learning
