Dimensionality Reduction Considered Harmful (Some of the Time)

TL;DR

This paper examines the reliability issues caused by dimensionality reduction in visual analytics and proposes new evaluation metrics, optimization strategies, and interaction techniques to improve trustworthiness.

Contribution

It introduces novel methods and metrics to assess and enhance the reliability of dimensionality reduction in visual analytics, addressing a critical challenge in the field.

Findings

New evaluation metrics for DR reliability

Optimization strategies to reduce DR errors

Interaction techniques to improve user trust

Abstract

Visual analytics now plays a central role in decision-making across diverse disciplines, but it can be unreliable: the knowledge or insights derived from the analysis may not accurately reflect the underlying data. In this dissertation, we improve the reliability of visual analytics with a focus on dimensionality reduction (DR). DR techniques enable visual analysis of high-dimensional data by reducing it to two or three dimensions, but they inherently introduce errors that can compromise the reliability of visual analytics. To this end, I investigate reliability challenges that practitioners face when using DR for visual analytics. Then, I propose technical solutions to address these challenges, including new evaluation metrics, optimization strategies, and interaction techniques. We conclude the thesis by discussing how our contributions lay the foundation for achieving more reliable visual analytics practices.