Assisted decision making using multivariate spatiotemporal data through the application of visual analytics
Abstract
The increasing availability of digital data provide both opportunities and challenges to analysts and decision makers. Data can be utilized to extract actionable information in a time constrained environment for making effective decisions. In order to better facilitate the exploration of such datasets, advanced tool sets are required that allow users to interact and provide insight into their data. In this work, we present a suite of visual analytic tools for spatiotemporal data exploration and analysis that enable analysts to make effective decisions and generate and test hypotheses pertaining to their datasets. These tools provide analysts with the ability to discover patterns and trends, allowing them to look for correlations and explore possible predictive links from among their datasets. Our research also applies a visual analytics approach for risk-based decision making scenarios in the spatiotemporal domain in order to assist users in understanding the implications of implementing different strategies. We also provide a visual analytics method for exploring correlations among multivariate datasets at different spatiotemporal scales, and present results that explore the benefits and harm of utilizing aggregate statistics. Finally, we present our proactive and predictive spatiotemporal visual analytics environment that enables decision makers to utilize their domain expertise at natural problem scales in order to help alleviate the cognitive overload caused due to the complexity of analysis process.
Degree
Ph.D.
Advisors
Ebert, Purdue University.
Subject Area
Computer Engineering
Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server.