Šourek, Gustav and Aschenbrenner, Vojtěch and Zelezný, Filip and Kuželka, Ondřej

Šourek, G., Aschenbrenner, V., Zelezný, F., & Kuželka, O. (2015). Lifted Relational Neural Networks. In Proceedings of the 2015th International Conference on Cognitive Computation: Integrating Neural and Symbolic Approaches - Volume 1583 (pp. 52–60). Aachen, DEU: CEUR-WS.org.

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 descent algorithm. The framework allows for hierarchical relational modeling constructs and learning of latent relational concepts through shared hidden layers weights corresponding to the rules. Discovery of notable relational concepts and experiments on 78 relational learning benchmarks demonstrate favorable performance of the method.

Citation

@inproceedings{lrnns2015,
  author = {Šourek, Gustav and Aschenbrenner, Vojtěch and Zelezný, Filip and Kuželka, Ondřej},
  title = {Lifted Relational Neural Networks},
  year = {2015},
  publisher = {CEUR-WS.org},
  address = {Aachen, DEU},
  booktitle = {Proceedings of the 2015th International Conference on Cognitive Computation: Integrating Neural and Symbolic Approaches - Volume 1583},
  pages = {52–60},
  numpages = {9},
  location = {Montreal, Canada},
  series = {COCO’15},
  url = {https://dl.acm.org/doi/abs/10.5555/2996831.2996838}
}