@inproceedings{Russo:2023:10.3233/faia230495,
author = {Russo, F and Toni, F},
doi = {10.3233/faia230495},
pages = {2025--2032},
publisher = {IOS Press},
title = {Causal discovery and knowledge injection for contestable neural networks},
url = {http://dx.doi.org/10.3233/faia230495},
year = {2023}
}

Download

TY  - CPAPER
AB  - Neural networks have proven to be effective at solvingmachine learning tasks but it is unclear whether they learn any relevant causal relationships, while their black-box nature makes it difficult for modellers to understand and debug them. We propose a novelmethod overcoming these issues by allowing a two-way interactionwhereby neural-network-empowered machines can expose the underpinning learnt causal graphs and humans can contest the machinesby modifying the causal graphs before re-injecting them into the machines, so that the learnt models are guaranteed to conform to thegraphs and adhere to expert knowledge (some of which can also begiven up-front). By building a window into the model behaviour andenabling knowledge injection, our method allows practitioners to debug networks based on the causal structure discovered from the dataand underpinning the predictions. Experiments with real and synthetic tabular data show that our method improves predictive performance up to 2.4x while producing parsimonious networks, up to 7xsmaller in the input layer, compared to SOTA regularised networks.
AU  - Russo,F
AU  - Toni,F
DO  - 10.3233/faia230495
EP  - 2032
PB  - IOS Press
PY  - 2023///
SN  - 0922-6389
SP  - 2025
TI  - Causal discovery and knowledge injection for contestable neural networks
UR  - http://dx.doi.org/10.3233/faia230495
UR  - http://hdl.handle.net/10044/1/105659
ER  - 

Download