Specialized Neural Network Pruning for Boolean Abstractions

Jarren Briscoe, Jarren Briscoe, Brian Rague, Kyle Feuz, Robert Ball

2021

Abstract

The inherent intricate topology of a neural network (NN) decreases our understanding of its function and purpose. Neural network abstraction and analysis techniques are designed to increase the comprehensibility of these computing structures. To achieve a more concise and interpretable representation of a NN as a Boolean graph (BG), we introduce the Neural Constantness Heuristic (NCH), Neural Constant Propagation (NCP), shared logic, the Neural Real-Valued Constantness Heuristic (NRVCH), and negligible neural nodes. These techniques reduce a neural layer’s input space and the number of nodes for a problem in NP (reducing its complexity). Additionally, we contrast two parsing methods that translate NNs to BGs: reverse traversal (N ) and forward traversal (F ). For most use cases, the combination of NRVCH, NCP, and N is the best choice.

Paper Citation

in Harvard Style

Briscoe J., Rague B., Feuz K. and Ball R. (2021). Specialized Neural Network Pruning for Boolean Abstractions. In Proceedings of the 13th International Joint Conference on Knowledge Discovery, Knowledge Engineering and Knowledge Management (IC3K 2021) - Volume 2: KEOD; ISBN 978-989-758-533-3, SciTePress, pages 178-185. DOI: 10.5220/0010657800003064

in Bibtex Style

@conference{keod21,
author={Jarren Briscoe and Brian Rague and Kyle Feuz and Robert Ball},
title={Specialized Neural Network Pruning for Boolean Abstractions},
booktitle={Proceedings of the 13th International Joint Conference on Knowledge Discovery, Knowledge Engineering and Knowledge Management (IC3K 2021) - Volume 2: KEOD},
year={2021},
pages={178-185},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0010657800003064},
isbn={978-989-758-533-3},
}

in EndNote Style

TY - CONF

JO - Proceedings of the 13th International Joint Conference on Knowledge Discovery, Knowledge Engineering and Knowledge Management (IC3K 2021) - Volume 2: KEOD
TI - Specialized Neural Network Pruning for Boolean Abstractions
SN - 978-989-758-533-3
AU - Briscoe J.
AU - Rague B.
AU - Feuz K.
AU - Ball R.
PY - 2021
SP - 178
EP - 185
DO - 10.5220/0010657800003064
PB - SciTePress