A Universal Logic Operator for Interpretable Deep Convolution Networks
Machine Learning
2019-01-25 v1 Machine Learning
Abstract
Explaining neural network computation in terms of probabilistic/fuzzy logical operations has attracted much attention due to its simplicity and high interpretability. Different choices of logical operators such as AND, OR and XOR give rise to another dimension for network optimization, and in this paper, we study the open problem of learning a universal logical operator without prescribing to any logical operations manually. Insightful observations along this exploration furnish deep convolution networks with a novel logical interpretation.
Cite
@article{arxiv.1901.08551,
title = {A Universal Logic Operator for Interpretable Deep Convolution Networks},
author = {KamWoh Ng and Lixin Fan and Chee Seng Chan},
journal= {arXiv preprint arXiv:1901.08551},
year = {2019}
}
Comments
In AAAI-19 Workshop on Network Interpretability for Deep Learning