Summary of On the Tractability Of Shap Explanations Under Markovian Distributions, by Reda Marzouk and Colin De La Higuera
On the Tractability of SHAP Explanations under Markovian Distributions
by Reda Marzouk, Colin de La Higuera
First submitted to arxiv on: 5 May 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 The SHAP framework is a widely used method for explaining machine learning models’ decisions. While it’s effective, computing its exact values can be challenging and even NP-Hard in some cases. Recent studies have made progress on specific model types, but these findings rely on the unrealistic assumption of independent features. This paper breaks new ground by introducing a Markovian perspective to relax this assumption and compute SHAP scores for certain model families in polynomial time. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary The SHAP framework is used to explain how machine learning models make predictions. It’s hard to calculate exactly, but some types of models can be easier. Researchers have found ways to do it more efficiently by assuming that features are not connected. This paper uses a different approach to solve the problem for certain kinds of models. |
Keywords
» Artificial intelligence » Machine learning
