Holographic Graph Neuron: a Bio-Inspired Architecture for Pattern Processing

TL;DR

This paper introduces a bio-inspired holographic graph neuron architecture utilizing Vector Symbolic Architectures, enabling efficient pattern memorization, noise resistance, and linear-time sub-pattern search.

Contribution

It presents a novel one-layered Hierarchical Graph Neuron architecture with improved noise resistance and efficient sub-pattern search capabilities using Vector Symbolic Architectures.

Findings

Maintains properties and performance of previous Hierarchical Graph Neuron models

Enhances noise resistance of the architecture

Achieves linear-time search for sub-patterns

Abstract

This article proposes the use of Vector Symbolic Architectures for implementing Hierarchical Graph Neuron, an architecture for memorizing patterns of generic sensor stimuli. The adoption of a Vector Symbolic representation ensures a one-layered design for the approach, while maintaining the previously reported properties and performance characteristics of Hierarchical Graph Neuron, and also improving the noise resistance of the architecture. The proposed architecture enables a linear (with respect to the number of stored entries) time search for an arbitrary sub-pattern.