Layerwise Learning of Mixed Conjunctive and Disjunctive Rule Networks
Sprache des Vortragstitels:
Englisch
Original Tagungtitel:
Spring workshop on Mining and Learning (SMiLe) 2023
Sprache des Tagungstitel:
Englisch
Original Kurzfassung:
Conventional rule learning algorithms learn a description of the positive class in disjunctive normal form (DNF). Alternatively, there are also a few learners who can formulate their model in conjunctive normal form (CNF) instead. While it is clear that both representations are equally expressive, there are domains where DNF learners perform better and others where CNF learners perform better. In this work, we propose a framework for learning general logical functions by training alternating layers of conjunctive and disjunctive rule sets, using any conventional rule learner. We evaluate its instantiation with LORD, a state-of-the-art propositional rule learner, which aims for efficiently learning the best rule for every training example.
Forschungs-