1 paper · 1 filter
Marco Pop-Mihali, Adrian Groza
We aim at development white-box machine learning algorithms. We focus here on algorithms for learning axioms in description logic. We extend the Class Expression Learning for Ontol…