Lifted Relational Neural Networks

Gustav Sourek, Vojtech Aschenbrenner, Filip Zelezný, Ondrej Kuzelka

Abstract: We propose a method combining relational-logic representations with neural network learning. A general lifted architecture, possibly reflecting some background domain knowledge, is described through relational rules which may be handcrafted or learned. The relational rule-set serves as a template for unfolding possibly deep neural networks whose structures also reflect the structures of given training or testing relational examples. Different networks corresponding to different examples share their weights, which co-evolve during training by stochastic gradient descend algorithm. Discovery of notable latent relational concepts and experiments on 78 relational learning benchmarks demonstrate favorable performance of the method.