Holographic Graph Neuron: a Bio-Inspired Architecture for Pattern Processing
This work addresses pattern processing challenges in bio-inspired computing, but it appears incremental as it builds on the existing Hierarchical Graph Neuron architecture.
The paper tackles the problem of memorizing patterns from generic sensor stimuli by proposing a Holographic Graph Neuron architecture using Vector Symbolic Architectures, resulting in a one-layered design with improved noise resistance and enabling linear-time search for arbitrary sub-patterns.
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.