Learning Higher-Order Programs Through Predicate Invention

Andrew Cropper, Rolf Morel, Stephen H. Muggleton

AAAI 2020 pp. 13655-13658

doi:10.1609/AAAI.V34I09.7113 /aaai/2020/cropper2020aaai-learning/

Abstract

A key feature of inductive logic programming (ILP) is its ability to learn first-order programs, which are intrinsically more expressive than propositional programs. In this paper, we introduce ILP techniques to learn higher-order programs. We implement our idea in Metagolho, an ILP system which can learn higher-order programs with higher-order predicate invention. Our experiments show that, compared to first-order programs, learning higher-order programs can significantly improve predictive accuracies and reduce learning times.

PDF AAAI Semantic Scholar

Cite

Text

Cropper et al. "Learning Higher-Order Programs Through Predicate Invention." AAAI Conference on Artificial Intelligence, 2020. doi:10.1609/AAAI.V34I09.7113

Markdown

[Cropper et al. "Learning Higher-Order Programs Through Predicate Invention." AAAI Conference on Artificial Intelligence, 2020.](https://mlanthology.org/aaai/2020/cropper2020aaai-learning/) doi:10.1609/AAAI.V34I09.7113

BibTeX

@inproceedings{cropper2020aaai-learning,
  title     = {{Learning Higher-Order Programs Through Predicate Invention}},
  author    = {Cropper, Andrew and Morel, Rolf and Muggleton, Stephen H.},
  booktitle = {AAAI Conference on Artificial Intelligence},
  year      = {2020},
  pages     = {13655-13658},
  doi       = {10.1609/AAAI.V34I09.7113},
  url       = {https://mlanthology.org/aaai/2020/cropper2020aaai-learning/}
}