A Visual Comparative Study on Multivariate Data Analysis

Dong, Yu

A Visual Comparative Study on Multivariate Data Analysis

Dong, Yu

Permalink

Publication Type:: Thesis
Issue Date:: 2023

Open Access

Copyright Clearance Process

Recently Added
In Progress
Open Access

This item is open access.

Adobe PDF

Download thesisAdobe PDF (9.87 MB)

View statistics

Full metadata record

Field	Value	Language
dc.contributor.author	Dong, Yu
dc.date.accessioned	2023-09-27T04:55:43Z
dc.date.available	2023-09-27T04:55:43Z
dc.date.issued	2023
dc.identifier.uri	http://hdl.handle.net/10453/172330
dc.description	University of Technology Sydney. Faculty of Engineering and Information Technology.	en_US.UTF-8
dc.description.abstract	Comparative analysis is crucial in real-world data analysis, especially for multivariate data. However, efficiently comparing hierarchies and node attributes, spatiotemporal data, and data uncertainties in multivariate data remains a challenge in certain applications. This thesis addresses this gap by providing a comprehensive visual comparative analysis of multivariate data. It introduces four research approaches that propose innovative visual solutions to enhance the understanding and comparison of such data. The first approach focuses on balancing the comparison of hierarchies and node attributes. It introduces a novel visualization technique called PansyTree, which utilizes a tree metaphor and node merging. By representing merged nodes in the structure, PansyTree enables the merging of three datasets into a single tree. This facilitates the exploration and comparison of structures, nodes, and node attributes. The second approach, called +msRNAer, presents a portrait-based visual modeling method for comparing time series and multidimensional features in epidemiology. It represents time series and multidimensional features in a reduced-dimensional space, creating portraits that highlight similarities and differences. This approach has been tested on COVID-19-related datasets and has proven effective in identifying location-based patterns and relationships between COVID-19 cases and risk factors in census data. The third approach, User-centered Visual Explorer (UcVE), offers customizable views for exploring and comparing spatiotemporal and multidimensional features. UcVE aims to reduce the cognitive load of users by providing visualization, saving, and tracking capabilities for exploration results. With its user-friendly interface, UcVE allows users to switch between views and explore data at different levels of detail, making the analysis of complex spatiotemporal data more accessible and intuitive. The fourth approach, ClinicLens, is an interactive visual analytics system designed to explore, compare, and optimize the testing capacities of healthcare clinics in the presence of multivariate uncertainties, specifically in the context of COVID-19. ClinicLens combines collaborative visual views with AI algorithms to assist domain experts in making informed decisions and adjustments regarding testing capacities and COVID-19 situations. In conclusion, these four approaches offer diverse perspectives for improving the understanding and comparison of complex multivariate data. They have demonstrated their usefulness and effectiveness in real-world applications, providing valuable insights and support for decision-making in various domains.	en_US.UTF-8
dc.format	Thesis (PhD)
dc.language.iso	en_US	en_US.UTF-8
dc.relation	https://opus.lib.uts.edu.au/bitstream/10453/172330/1/thesis.pdf
dc.rights	info:eu-repo/semantics/openAccess
dc.rights	The author owns the copyright in this thesis including all reproduction and reuse rights for the work. The work may not be altered without the permission of the copyright owner. Attribution is essential when quoting or paraphrasing from this thesis.
dc.rights	© 2023 Yu Dong
dc.rights	au.edu.uts.lib/cph
dc.title	A Visual Comparative Study on Multivariate Data Analysis	en_US.UTF-8
dc.type	Thesis
utslib.copyright.status	open_access	*

Abstract:

Comparative analysis is crucial in real-world data analysis, especially for multivariate data. However, efficiently comparing hierarchies and node attributes, spatiotemporal data, and data uncertainties in multivariate data remains a challenge in certain applications. This thesis addresses this gap by providing a comprehensive visual comparative analysis of multivariate data. It introduces four research approaches that propose innovative visual solutions to enhance the understanding and comparison of such data. The first approach focuses on balancing the comparison of hierarchies and node attributes. It introduces a novel visualization technique called PansyTree, which utilizes a tree metaphor and node merging. By representing merged nodes in the structure, PansyTree enables the merging of three datasets into a single tree. This facilitates the exploration and comparison of structures, nodes, and node attributes. The second approach, called +msRNAer, presents a portrait-based visual modeling method for comparing time series and multidimensional features in epidemiology. It represents time series and multidimensional features in a reduced-dimensional space, creating portraits that highlight similarities and differences. This approach has been tested on COVID-19-related datasets and has proven effective in identifying location-based patterns and relationships between COVID-19 cases and risk factors in census data. The third approach, User-centered Visual Explorer (UcVE), offers customizable views for exploring and comparing spatiotemporal and multidimensional features. UcVE aims to reduce the cognitive load of users by providing visualization, saving, and tracking capabilities for exploration results. With its user-friendly interface, UcVE allows users to switch between views and explore data at different levels of detail, making the analysis of complex spatiotemporal data more accessible and intuitive. The fourth approach, ClinicLens, is an interactive visual analytics system designed to explore, compare, and optimize the testing capacities of healthcare clinics in the presence of multivariate uncertainties, specifically in the context of COVID-19. ClinicLens combines collaborative visual views with AI algorithms to assist domain experts in making informed decisions and adjustments regarding testing capacities and COVID-19 situations. In conclusion, these four approaches offer diverse perspectives for improving the understanding and comparison of complex multivariate data. They have demonstrated their usefulness and effectiveness in real-world applications, providing valuable insights and support for decision-making in various domains.

Please use this identifier to cite or link to this item:

http://hdl.handle.net/10453/172330