Theoretical analysis of dynamic binary neural networks with simple sparse connection

Aoki, Sadao; Koyama, Seitaro; Saito, Toshimichi

doi:10.1016/j.neucom.2019.03.015

Cited by 12 publications

(3 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The PBNN can be regarded as a simplified version of the dynamic binary neural network (DBNN [3] [11] [22]) characterized by global ternary connection. Depending on the parameters and initial conditions, the PBNN and DBNN generate various periodic orbits of binary vectors (BPOs).…”

mentioning

confidence: 99%

See 1 more Smart Citation

Permutation binary neural networks: Analysis of periodic orbits and its applications

Udagawa

Okano

Saito

2023

DCDS-B

Self Cite

View full text Add to dashboard Cite

<p style='text-indent:20px;'>This paper presents a permutation binary neural network characterized by local binary connection, global permutation connection, and the signum activation function. The dynamics is described by a difference equation of binary state variables. Depending on the connection, the network generates various periodic orbits of binary vectors. The binary/permutation connection brings benefits to precise analysis and to FPGA based hardware implementation. In order to consider the periodic orbits, we introduce three tools: a composition return map for visualization of the dynamics, two feature quantities for classification of periodic orbits, and an FPGA based hardware prototype for engineering applications. Using the tools, we have analyzed all the 6-dimensional networks. Typical periodic orbits are confirmed experimentally.</p>

show abstract

mentioning

confidence: 99%

“…In [22], we have considered storage and stabilization of special BPOs for application to switching power converters. In [3], we have analyzed the DBNNs with special sparse connection and have given theoretical results for storage and stabilization of rotation-type BPOs. However, in these DBNNs, storage of any BPO is not guaranteed.…”

mentioning

confidence: 99%

Permutation binary neural networks: Analysis of periodic orbits and its applications

Udagawa

Okano

Saito

2023

DCDS-B

Self Cite

View full text Add to dashboard Cite

show abstract

“…In [9], we have considered storage and stabilization of special BPOs for application to switching power converters. In [10], we have analyzed the DBNNs with special sparse connection and have given theoretical results for storage and stabilization of rotation-type BPOs. However, in these DBNNs, storage of any BPO is not guaranteed.…”

mentioning

confidence: 99%

Permutation Binary Neural Networks: Analysis of Periodic Orbits and Its Applications

Udagawa¹,

Okano²,

Saito³

2022

Preprint

Self Cite

View full text Add to dashboard Cite

This paper presents a permutation binary neural network characterized by local binary connection, global permutation connection, and the signum activation function. The dynamics is described by a difference equation of binary state variables. Depending on the connection, the network generates various periodic orbits of binary vectors. The binary/permutation connection brings benefits to precise analysis and to FPGA based hardware implementation. In order to consider the periodic orbits, we introduce three tools: a composition return map for visualization of the dynamics, two feature quantities for classification of periodic orbits, and an FPGA based hardware prototype for engineering applications. Using the tools, we have analyzed all the 6-dimensional networks. Typical periodic orbits are confirmed experimentally.

show abstract

Sparse Dynamic Binary Neural Networks for Storage and Switching of Binary Periodic Orbits

Anzai

Koyama

Aoki

et al. 2019

Neural Information Processing

View full text Add to dashboard Cite

Theoretical analysis of dynamic binary neural networks with simple sparse connection

Cited by 12 publications

References 17 publications

Permutation binary neural networks: Analysis of periodic orbits and its applications

Permutation binary neural networks: Analysis of periodic orbits and its applications

Permutation Binary Neural Networks: Analysis of Periodic Orbits and Its Applications

Sparse Dynamic Binary Neural Networks for Storage and Switching of Binary Periodic Orbits

Contact Info

Product

Resources

About