Please use this identifier to cite or link to this item: https://hdl.handle.net/10356/160695
Title: Taylor's theorem: a new perspective for neural tensor networks
Authors: Li, Wei
Zhu, Luyao
Cambria, Erik
Keywords: Engineering::Computer science and engineering
Issue Date: 2021
Source: Li, W., Zhu, L. & Cambria, E. (2021). Taylor's theorem: a new perspective for neural tensor networks. Knowledge-Based Systems, 228, 107258-. https://dx.doi.org/10.1016/j.knosys.2021.107258
Project: A18A2b0046
Journal: Knowledge-Based Systems
Abstract: Neural tensor networks have been widely used in a large number of natural language processing tasks such as conversational sentiment analysis, named entity recognition and knowledge base completion. However, the mathematical explanation of neural tensor networks remains a challenging problem, due to the bilinear term. According to Taylor's theorem, a kth order differentiable function can be approximated by a kth order Taylor polynomial around a given point. Therefore, we provide a mathematical explanation of neural tensor networks and also reveal the inner link between them and feedforward neural networks from the perspective of Taylor's theorem. In addition, we unify two forms of neural tensor networks into a single framework and present factorization methods to make the neural tensor networks parameter-efficient. Experimental results bring some valuable insights into neural tensor networks.
URI: https://hdl.handle.net/10356/160695
ISSN: 0950-7051
DOI: 10.1016/j.knosys.2021.107258
Rights: © 2021 Elsevier B.V. All rights reserved.
Fulltext Permission: none
Fulltext Availability: No Fulltext
Appears in Collections:SCSE Journal Articles

SCOPUSTM   
Citations 20

12
Updated on Jan 28, 2023

Web of ScienceTM
Citations 20

10
Updated on Jan 29, 2023

Page view(s)

22
Updated on Jan 30, 2023

Google ScholarTM

Check

Altmetric


Plumx

Items in DR-NTU are protected by copyright, with all rights reserved, unless otherwise indicated.