Visualization of Manifold-Valued Elements by Multidimensional Scaling

TL;DR

This paper proposes using multidimensional scaling (MDS) to visualize elements on high-dimensional manifolds, aiding understanding in signal processing and machine learning contexts.

Contribution

It introduces a novel application of MDS as a visualization tool specifically for manifold-valued elements in high-dimensional spaces.

Findings

MDS effectively visualizes high-dimensional manifold elements.

The approach benefits signal processing and machine learning analysis.

Visualization improves understanding of complex parameter spaces.

Abstract

The present contribution suggests the use of a multidimensional scaling (MDS) algorithm as a visualization tool for manifold-valued elements. A visualization tool of this kind is useful in signal processing and machine learning whenever learning/adaptation algorithms insist on high-dimensional parameter manifolds.