https://github.com/amazon-science/explaining-probabilistic-models-with-distributinal-values
Distributional values for XAI
https://github.com/amazon-science/explaining-probabilistic-models-with-distributinal-values
Science Score: 23.0%
This score indicates how likely this project is to be science-related based on various indicators:
-
○CITATION.cff file
-
✓codemeta.json file
Found codemeta.json file -
○.zenodo.json file
-
○DOI references
-
✓Academic publication links
Links to: arxiv.org -
○Academic email domains
-
○Institutional organization owner
-
○JOSS paper metadata
-
○Scientific vocabulary similarity
Low similarity (11.0%) to scientific vocabulary
Repository
Distributional values for XAI
Basic Info
Statistics
- Stars: 6
- Watchers: 1
- Forks: 0
- Open Issues: 0
- Releases: 0
Metadata Files
README.md
Explaining Probabilistic Models with Distributional Values
Implementation of the distributional values framework for XAI
Distributional values operators extend the classic value operators (like the Shapley value) in cooperative game theory to process cooperative game with stochastic payoffs.
A distributional value of a player of a cooperative stochastic game is a random variable representing the contribution of that player toward achieving the grand payoff, i.e. the payoff attained when all players play the game.
In the context of explainability, one may derive stochastic games from probabilistic (binary or multiclass) classifiers, like neural nets. The common usage in XAI is to identify players with input features and compute attributions for such features. In this context, distributional values can help answer questions such as: - ''what is the probability that a feature does not contribute in changing prediction?'' - ''what is the feature that impacted the most for making $x$ being classified as an airplane rather than a dog?'' - ''what is the variance of the contribution of the input features?''.
Furthermore, the expectation of a distributional value yields a standard value, linking the two concepts closely and making the distributional values a strict extension of the standard one.
Quickstart
Example of the Bernoulli Shapley value for a XOR game
Other examples
- Categorical Shapley value for Iris multinomial regression
- Categorical Shapley value for Mnist (explaining the 8)
Please open an issue should you be interested in code covering the other case studies in the paper.
Paper
@article{franceschi2024explaining,
title={Explaining Probabilistic Models with Distributional Values},
author={Franceschi, Luca and Donini, Michele and Archambeau, C{\'e}dric and Seeger, Matthias},
journal={ICML},
year={2024}
}
License
Please see LICENSE
Notice
This code is being released solely for academic and scientific reproducibility purposes, in support of the methods and findings described in the associated publication. Pull requests are not being accepted in order to maintain the code exactly as it was used in the paper, but interested parties are encouraged to open an issue requesting open source community development.
Owner
- Name: Amazon Science
- Login: amazon-science
- Kind: organization
- Website: https://amazon.science
- Twitter: AmazonScience
- Repositories: 80
- Profile: https://github.com/amazon-science
GitHub Events
Total
- Watch event: 1
Last Year
- Watch event: 1
Issues and Pull Requests
Last synced: over 1 year ago
All Time
- Total issues: 0
- Total pull requests: 0
- Average time to close issues: N/A
- Average time to close pull requests: N/A
- Total issue authors: 0
- Total pull request authors: 0
- Average comments per issue: 0
- Average comments per pull request: 0
- Merged pull requests: 0
- Bot issues: 0
- Bot pull requests: 0
Past Year
- Issues: 0
- Pull requests: 0
- Average time to close issues: N/A
- Average time to close pull requests: N/A
- Issue authors: 0
- Pull request authors: 0
- Average comments per issue: 0
- Average comments per pull request: 0
- Merged pull requests: 0
- Bot issues: 0
- Bot pull requests: 0
Top Authors
Issue Authors
Pull Request Authors
Top Labels
Issue Labels
Pull Request Labels
Dependencies
- jupyter *
- matplotlib *
- numpy *
- pyfunctional *
- scikit-learn *
- torch *
- torchvision *
- tqdm *