README.md

Explainability in Practice

Code for the paper "Explainability in Practice: Estimating Electrification Rates from Mobile Phone Data in Senegal" (L. State, H. Salat, S. Rubrichi and Z. Smoreda)

Version 2 (newer version):

Archival at The first World Conference on eXplainable AI (XAI 2023)

Updated code (and paper).

Version 1 (older version):

Non-archival at TSRML Workshop (NeurIPS 2022)

Jupyter notebook files:

1. training the classifiers

2. generating the explanations for LIME and SHAP separately, 3 different notebooks as LIME generation is separated from plotting

You can find the paper here