Transparent Explainable Logic Layers

Alessio Ragno, Marc Plantevit, Celine Robardet and Roberto Capobianco

Published Paper: https://doi.org/10.3233/FAIA240579

Official repository of the ECAI2024 Paper. The code is based on the LENs repository: https://github.com/pietrobarbiero/logic_explained_networks.git

The code to reproduce the experiments in the paper is in the experiments/ folder. Some datasets must be downloaded from the original websites:

• MIMIC: https://www.physionet.org/content/mimic2-iaccd/1.0/full_cohort_data.csv

The full implementation of TELL is in lens/models/logic.py.

A lighter implementation paired with an example is available in folder example/

Cite our work:

@inbook{TELL2024,
  title = {Transparent Explainable Logic Layers},
  ISBN = {9781643685489},
  ISSN = {1879-8314},
  url = {http://dx.doi.org/10.3233/FAIA240579},
  DOI = {10.3233/faia240579},
  booktitle = {ECAI 2024},
  publisher = {IOS Press},
  author = {Ragno,  Alessio and Plantevit,  Marc and Robardet,  Celine and Capobianco,  Roberto},
  year = {2024},
  month = oct 
}