English

Weight Priors for Learning Identity Relations

Machine Learning 2020-07-07 v2 Machine Learning

Abstract

Learning abstract and systematic relations has been an open issue in neural network learning for over 30 years. It has been shown recently that neural networks do not learn relations based on identity and are unable to generalize well to unseen data. The Relation Based Pattern (RBP) approach has been proposed as a solution for this problem. In this work, we extend RBP by realizing it as a Bayesian prior on network weights to model the identity relations. This weight prior leads to a modified regularization term in otherwise standard network learning. In our experiments, we show that the Bayesian weight priors lead to perfect generalization when learning identity based relations and do not impede general neural network learning. We believe that the approach of creating an inductive bias with weight priors can be extended easily to other forms of relations and will be beneficial for many other learning tasks.

Keywords

Cite

@article{arxiv.2003.03125,
  title  = {Weight Priors for Learning Identity Relations},
  author = {Radha Kopparti and Tillman Weyde},
  journal= {arXiv preprint arXiv:2003.03125},
  year   = {2020}
}

Comments

Proceedings of KR2ML @ NeurIPS 2019, Vancouver, Canada

R2 v1 2026-06-23T14:06:19.085Z