Holographic Graph Neuron: A Bioinspired Architecture for Pattern Processing

Holographic Graph Neuron: A Bioinspired Architecture for Pattern Processing
复制标题

全息图神经元:用于模式处理的仿生架构

DOI:
--
复制
发表时间:
2015
影响因子:
10.4
通讯作者:
Y. Sekercioglu
Y. Sekercioglu
中科院分区:
计算机科学1区
文献类型:
--
作者:
D. Kleyko;Evgeny Osipov;A. Senior;Asad I. Khan;Y. Sekercioglu

文献摘要

被引文献

相似文献

在本文中,我们提出了一种新的方法来实现层次图神经元(HGN),一个架构,用于记忆模式的通用传感器刺激,通过使用矢量符号架构。采用矢量符号表示确保了单层设计,同时保留了HGN的现有性能特征。这种方法显着提高了HGN架构的抗噪声能力,并实现了对任意子模式的线性(相对于存储条目的数量)时间搜索。
In this paper, we propose a new approach to implementing hierarchical graph neuron (HGN), an architecture for memorizing patterns of generic sensor stimuli, through the use of vector symbolic architectures. The adoption of a vector symbolic representation ensures a single-layer design while retaining the existing performance characteristics of HGN. This approach significantly improves the noise resistance of the HGN architecture, and enables a linear (with respect to the number of stored entries) time search for an arbitrary subpattern.