Learning Logistic Circuits

AAAI 2019 pp. 4277-4286

doi:10.1609/AAAI.V33I01.33014277 /aaai/2019/liang2019aaai-learning/

Abstract

This paper proposes a new classification model called logistic circuits. On MNIST and Fashion datasets, our learning algorithm outperforms neural networks that have an order of magnitude more parameters. Yet, logistic circuits have a distinct origin in symbolic AI, forming a discriminative counterpart to probabilistic-logical circuits such as ACs, SPNs, and PSDDs. We show that parameter learning for logistic circuits is convex optimization, and that a simple local search algorithm can induce strong model structures from data.

PDF AAAI Semantic Scholar

Cite

Text

Liang and Van den Broeck. "Learning Logistic Circuits." AAAI Conference on Artificial Intelligence, 2019. doi:10.1609/AAAI.V33I01.33014277

Markdown

[Liang and Van den Broeck. "Learning Logistic Circuits." AAAI Conference on Artificial Intelligence, 2019.](https://mlanthology.org/aaai/2019/liang2019aaai-learning/) doi:10.1609/AAAI.V33I01.33014277

BibTeX

@inproceedings{liang2019aaai-learning,
  title     = {{Learning Logistic Circuits}},
  author    = {Liang, Yitao and Van den Broeck, Guy},
  booktitle = {AAAI Conference on Artificial Intelligence},
  year      = {2019},
  pages     = {4277-4286},
  doi       = {10.1609/AAAI.V33I01.33014277},
  url       = {https://mlanthology.org/aaai/2019/liang2019aaai-learning/}
}