xai-regression
Science Score: 67.0%
This score indicates how likely this project is to be science-related based on various indicators:
-
✓CITATION.cff file
Found CITATION.cff file -
✓codemeta.json file
Found codemeta.json file -
✓.zenodo.json file
Found .zenodo.json file -
✓DOI references
Found 1 DOI reference(s) in README -
✓Academic publication links
Links to: arxiv.org, ieee.org -
○Academic email domains
-
○Institutional organization owner
-
○JOSS paper metadata
-
○Scientific vocabulary similarity
Low similarity (12.2%) to scientific vocabulary
Repository
Basic Info
- Host: GitHub
- Owner: sltzgs
- License: mit
- Language: Jupyter Notebook
- Default Branch: main
- Size: 544 KB
Statistics
- Stars: 27
- Watchers: 8
- Forks: 4
- Open Issues: 0
- Releases: 0
Metadata Files
README.md
Toward Explainable Artificial Intelligence for Regression Models: A methodological perspective
About
Welcome to xai-regression. We think of explainable artificial intelligence (XAI) as crucial for an informed application of machine learning methods in practice. Although XAI techniques have reached significant popularity for classifiers, little attention has been devoted to XAI for regression models so far. With our IEEE SPM publication Toward Explainable Artificial Intelligence for Regression Models: A methodological perspective (open pre-print available here) and the corresponding website we aim to address this gap and provide an overview of the latest developments of the field.
This repository contains implementations of the restructuring approach presented in the above mentioned paper.
If you are using the code, please cite it as:
sh
@ARTICLE{Letzgus_XAIR_2022,
author={Letzgus, Simon and Wagner, Patrick and Lederer, Jonas and Samek, Wojciech and Müller, Klaus-Robert and Montavon, Grégoire},
journal={IEEE Signal Processing Magazine},
title={Toward Explainable Artificial Intelligence for Regression Models: A methodological perspective},
year={2022},
volume={39},
number={4},
pages={40-58},
doi={10.1109/MSP.2022.3153277}}
Getting Started
The restructuring_implementation.ipynb contains functions implementing the proposed flooding-rule (findaref()) for latent offset selection and the introduced restructuring of ANN top-layers (restructure_model()) alongside the respective parts of the paper.
Prerequisites
The required libraries are listed in the requirements.txt-file.
Usage
The functions can be used to explain your model output relative to custom-specific reference values using existing propagation-based XAI techniques, such as LRP (pytorch-implementation available in the Zennit library).
Contact
simon.leszek@campus.tu-berlin.de
Machine Learning Group, Technische Universität Berlin, Straße des 17. Juni 135, Berlin, 10623, Germany.
Owner
- Name: Simon Letzgus
- Login: sltzgs
- Kind: user
- Location: Berlin
- Company: Technische Universität Berlin
- Repositories: 2
- Profile: https://github.com/sltzgs
Citation (CITATION.cff)
cff-version: 1.2.0
message: "If you use this software, please cite it as below."
authors:
- family-names: Letzgus, Wagner, Lederer, Samek, Müller, Montavon
given-names: Simon, Patrick, Jonas, Wojciech, Klaus-Robert, Gregoire
title: "Toward Explainable Artificial Intelligence for Regression Models: A methodological perspective"
version: 1.0.0
doi: 10.1109/MSP.2022.3153277
date-released: 2022-06-28
GitHub Events
Total
- Watch event: 3
- Push event: 1
Last Year
- Watch event: 3
- Push event: 1
Dependencies
- matplotlib ==3.5.3
- numpy ==1.21.5
- torch ==1.11.0