Keywords

Visual analytics; multi-objective optimization; Evolutionary algorithms

Location

Session D2: Virtual Reality, 3D Applications, and Immersive Visualization

Start Date

17-6-2014 2:00 PM

End Date

17-6-2014 3:20 PM

Abstract

In the US, the recently introduced Principles and Requirements for Federal Investment in Water Resources captures an increasing trend for using multiple objective approaches to make funding decisions for water resources and environmental problems. Moreover, advanced search tools such as multi objective evolutionary algorithms can develop solutions to such problems, even when the problems are characterized by non-linearities, uncertainty, and false optima. However, our presentation argues that multi-objective decision support requires innovative methods for visualizing complex trade-offs between conflicting objectives, interrogating design possibilities, imposing preferences between objectives, and eventually negotiating a small set of final alternatives. Specifically we present a recently developed version of a software framework called AeroVis, which interfaces with spreadsheet software to integrate tabular visualizations of data with immersive, three dimensional glyph plots that show up to 7 objectives simultaneously. AeroVis has been used both in the design of complex engineered systems as well as water resources management decision making problems. Our results show how a large amount of information can be quickly processed by the user, integrating the emerging concept of "big data" to support environmental decision making with human pattern recognition.

 
Jun 17th, 2:00 PM Jun 17th, 3:20 PM

Balancing Conflicting Management Objectives Using Interactive, Three-Dimensional Visual Analytics

Session D2: Virtual Reality, 3D Applications, and Immersive Visualization

In the US, the recently introduced Principles and Requirements for Federal Investment in Water Resources captures an increasing trend for using multiple objective approaches to make funding decisions for water resources and environmental problems. Moreover, advanced search tools such as multi objective evolutionary algorithms can develop solutions to such problems, even when the problems are characterized by non-linearities, uncertainty, and false optima. However, our presentation argues that multi-objective decision support requires innovative methods for visualizing complex trade-offs between conflicting objectives, interrogating design possibilities, imposing preferences between objectives, and eventually negotiating a small set of final alternatives. Specifically we present a recently developed version of a software framework called AeroVis, which interfaces with spreadsheet software to integrate tabular visualizations of data with immersive, three dimensional glyph plots that show up to 7 objectives simultaneously. AeroVis has been used both in the design of complex engineered systems as well as water resources management decision making problems. Our results show how a large amount of information can be quickly processed by the user, integrating the emerging concept of "big data" to support environmental decision making with human pattern recognition.