Unsupervised Learning in Neuromemristive Systems

TL;DR

This paper explores unsupervised clustering in neuromemristive systems, demonstrating a memristor-based approach that achieves performance comparable to traditional k-means clustering, advancing energy-efficient neuro-inspired computation.

Contribution

It introduces a memristor crossbar architecture for unsupervised clustering within NMSs, filling a research gap in neuro-inspired unsupervised learning methods.

Findings

Achieved clustering performance comparable to MATLAB's k-means.

Demonstrated the viability of simple memristor crossbar architectures for unsupervised learning.

Contributed to energy-efficient neuro-inspired computation methods.

Abstract

Neuromemristive systems (NMSs) currently represent the most promising platform to achieve energy efficient neuro-inspired computation. However, since the research field is less than a decade old, there are still countless algorithms and design paradigms to be explored within these systems. One particular domain that remains to be fully investigated within NMSs is unsupervised learning. In this work, we explore the design of an NMS for unsupervised clustering, which is a critical element of several machine learning algorithms. Using a simple memristor crossbar architecture and learning rule, we are able to achieve performance which is on par with MATLAB's k-means clustering.