Salvato in:
Dettagli Bibliografici
Autore principale: Leconte, Louis
Natura: Preprint
Pubblicazione: 2024
Soggetti:
Accesso online:https://arxiv.org/abs/2401.16418
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866929227317641216
author Leconte, Louis
author_facet Leconte, Louis
contents The notion of Boolean logic backpropagation was introduced to build neural networks with weights and activations being Boolean numbers. Most of computations can be done with Boolean logic instead of real arithmetic, both during training and inference phases. But the underlying discrete optimization problem is NP-hard, and the Boolean logic has no guarantee. In this work we propose the first convergence analysis, under standard non-convex assumptions.
format Preprint
id arxiv_https___arxiv_org_abs_2401_16418
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Boolean Logic as an Error feedback mechanism
Leconte, Louis
Machine Learning
The notion of Boolean logic backpropagation was introduced to build neural networks with weights and activations being Boolean numbers. Most of computations can be done with Boolean logic instead of real arithmetic, both during training and inference phases. But the underlying discrete optimization problem is NP-hard, and the Boolean logic has no guarantee. In this work we propose the first convergence analysis, under standard non-convex assumptions.
title Boolean Logic as an Error feedback mechanism
topic Machine Learning
url https://arxiv.org/abs/2401.16418