Fastest learning in small-world neural networks

2004/02/29 by D. Simard, Denys Simard, Line Nadeau +2
Computer Science · Physics and Astronomy · #Machine Learning and ELM #Neural Networks and Applications #Stochastic Gradient Optimization Techniques #cond-mat.dis-nn #physics.bio-ph

paper · pdf · doi:10.1016/j.physleta.2004.12.078

published as Phys. Lett. A336 (2005) 8-15. · Text completely revised (14 pages), all new figures (7 figs)

arxiv created 2005/01/06 · openalex publication_date 2005/01/11 · arxiv updated 2009/12/01 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28

Abstract

We investigate supervised learning in neural networks. We consider a multi-layered feed-forward network with back propagation. We find that the network of small-world connectivity reduces the learning error and learning time when compared to the networks of regular or random connectivity. Our study has potential applications in the domain of data-mining, image processing, speech recognition, and pattern recognition.

Citations

• Emergence of Scaling in Random Networks
• Error and attack tolerance of complex networks
• Gradient-based learning applied to document recognition
• The large-scale organization of metabolic networks
• On power-law relationships of the Internet topology