Princeton University Library Catalog

Beyond3D: Visualizing High-Dimensional Data Sets

Author/​Artist:
Durst, David [Browse]
Format:
Senior thesis
Language:
English
Advisor(s):
Braverman, Mark [Browse]
Contributor(s):
Department:
Princeton University. Department of Computer Science [Browse]
Class year:
2015
Description:
69 pages
Summary note:
I present Beyond3D, a system that explains high-dimensional data sets to non-technical users with domain-specific knowledge through intuitive, interactive visualizations. Beyond3D surpasses current systems which fail to handle large numbers of dimensions, like matrices of scatterplots and histograms, or require technical training, such as PCA. A combination of user testing and a case study demonstrates Beyond3D’s effectiveness. In the user study, non-technical test subjects with significant financial knowledge used the tool to answer questions about a financial data set. The case study demonstrates how a non-technical investor could use Beyond3D to decrease losses during market crashes.