Hierarchical Knowledge Extraction from Opaque Machine Learning Predictors

Adopting opaque machine learning predictors, which achieve
very high predictive performance, often necessitates incorporating sym-
bolic knowledge-extraction techniques. These techniques aim to explain
the opaque predictions, thus making them applicable in high-stakes sce-
narios. The development of symbolic knowledge-extraction procedures is
evolving alongside the dynamic machine learning landscape. However,
there are recurring drawbacks that tend to be overlooked or addressed in
a suboptimum way. Common examples include the non-exhaustiveness
of the global explanations generated for a black-box predictor or the
unwanted discretisation introduced in the prediction of continuous vari-
ables. To tackle these challenges, in this work, we introduce the HEx
algorithm, its formalisation and its properties. This algorithm aims to
obtain a symbolic, hierarchical representation of the knowledge acquired
by opaque machine learning classifiers and regressors, always ensuring
knowledge exhaustiveness and avoiding any output discretisation. Ex-
periments demonstrating the superior capabilities of HEx compared to
state-of-the-art competitors in terms of predictive performance, com-
pleteness, and human readability are presented.

keywords   Explainable artificial intelligence· Symbolic knowledge ex- traction· PSyKE

origin event

 AIxIA 2024

funding project

 AEQUITAS — Assessment and Engineering of eQuitable, Unbiased, Impartial and Trustworthy Ai Systems (01/11/2022–31/10/2025)

 FAIR-PE01-SP08 — Future AI Research – Partenariato Esteso sull'Intelligenza Artificiale – Spoke 8 “Pervasive AI” (01/01/2023–31/12/2025)