DSpace Repository

VISUAL ASSOCIATION MINING OF MULTIVARIATE DATA

Show simple item record

dc.contributor.advisor Mueller, Klaus en_US
dc.contributor.author Zhang, Zhiyuan en_US
dc.contributor.other Department of Computer Science. en_US
dc.date.accessioned 2017-09-20T16:52:31Z
dc.date.available 2017-09-20T16:52:31Z
dc.date.issued 2014-12-01 en_US
dc.identifier.uri http://hdl.handle.net/11401/77323 en_US
dc.description 112 pg. en_US
dc.description.abstract The rapid development of information technology produces vast amounts of data with numerous attributes. These multi-dimensional datasets offer tremendous opportunities for studying existing behavioral patterns and for predicting future developments. However, the high-dimensional space exceeds human comprehension. More sophisticated visualization techniques than the arsenal of standard plots are needed. First, we introduce an interactive navigation technique to help the analysts explore within the multi-dimensional data spaces. We employ a network-based interface and pair it with a parallel coordinates plot. In the network interface, the dimensions form nodes that are connected by edges representing the strength of association between dimensions. The analysts can interactively manipulate a route in the network, which is captured by the parallel coordinates plot in the form of the dimension ordering. Then, we extend the navigation interface to interactive correlation and causation analysis for both numerical and categorical variables within a unified framework. We also build a landscape (map) out of the network, which shows the raw data within the network and helps analysts quickly learn relationships and trends of the data. We demonstrate it via several applications, such as helping statisticians with model discovery. Furthermore, we prove the viability of our framework in the context of real scientific problem--climate research, and show how it helps a team of scientists make important discoveries. Finally, we introduce an interactive visual analytics interface designed for the healthcare informatics. It uses the Five-W's to establish a comprehensive multi-faceted assessment of the patient's history. The patient's multivariate data is visualized by associating each such W with a dedicated visual encoding that can represent and communicate it in effective ways. en_US
dc.description.sponsorship This work is sponsored by the Stony Brook University Graduate School in compliance with the requirements for completion of degree. en_US
dc.format Monograph en_US
dc.format.medium Electronic Resource en_US
dc.language.iso en_US en_US
dc.publisher The Graduate School, Stony Brook University: Stony Brook, NY. en_US
dc.subject.lcsh Computer science en_US
dc.subject.other Association Mining, Correlation Analysis, Healthcare, Information Visualization, Multivariate Data, Visual Analytics en_US
dc.title VISUAL ASSOCIATION MINING OF MULTIVARIATE DATA en_US
dc.type Dissertation en_US
dc.mimetype Application/PDF en_US
dc.contributor.committeemember Ramakrishnan, IV en_US
dc.contributor.committeemember Ortiz, Luis en_US
dc.contributor.committeemember McDonnell, Kevin. en_US


Files in this item

This item appears in the following Collection(s)

Show simple item record

Search DSpace


Advanced Search

Browse

My Account