.png){kind=icon}
.png){kind=icon}
Return
Collaboration pattern in parallel-connected neural networks – A potential criterion for measuring the bionic level of artificial neural networks
.png){kind=icon}
PRE
AI
G
DOI：10.1016/j.eswa.2026.133956.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Highlight • Sub-networks of an optimized parallel-connected neural network stay not optimized. • Inferior sub-networks contribute as much as superior sub-networks. • The connected network outputs correctly when all sub-networks give incorrect outputs. • The ReLU function behaves more bionic than the Tanh function does.
Keywords:
Artificial neural network
Parallel architecture
Activation function
Bionics
Supervised learning
Journal
IF:
7.5
Papers:
2.9W
Citations:
10.2W
Organization
No organization information available
Cited Papers