Large-Margin Hyperdimensional Computing: A Learning-Theoretical Perspective

TL;DR

This paper introduces a maximum-margin hyperdimensional computing classifier that outperforms existing methods, leveraging a novel formal relation to support vector machines to enable more efficient resource-constrained learning.

Contribution

It establishes a formal relation between hyperdimensional computing and SVMs and proposes a maximum-margin HDC classifier with improved performance.

Findings

Significant performance improvement over baseline HDC methods.

First formal relation established between HDC and SVMs.

Potential for more hardware-efficient learning solutions.

Abstract

Overparameterized machine learning (ML) methods such as neural networks may be prohibitively resource intensive for devices with limited computational capabilities. Hyperdimensional computing (HDC) is an emerging resource efficient and low-complexity ML method that allows hardware efficient implementations of (re-)training and inference procedures. In this paper, we propose a maximum-margin HDC classifier, which significantly outperforms baseline HDC methods on several benchmark datasets. Our method leverages a formal relation between HDC and support vector machines (SVMs) that we established for the first time. Our findings may inspire novel HDC methods with potentially more hardware-oriented implementations compared to SVMs, thus enabling more efficient learning solutions for various intelligent resource-constrained applications.