Visual Design Space Exploration using Contextual Self-Organizing Maps

Nekolny, Brett; Richardson, Trevor; Winer, Eliot H.

doi:10.2514/6.2010-9326

Cited by 6 publications

(2 citation statements)

References 4 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…One possibility for visualizing a many-dimensional pareto-front is the pairwise mapping to two-dimensional pareto-front curves, but this method is not acceptable for understandable and user-friendly presentation. A promising alternative is clustering using self-organizing maps (SOM) [16][17][18].…”

Section: Clustering Of Design Variantsmentioning

confidence: 99%

Urban Design Synthesis for Building Layouts Based on Evolutionary Many-Criteria Optimization

König

2015

International Journal of Architectural Computing

View full text Add to dashboard Cite

When working on urban planning projects there are usually multiple aspects to consider. Often these aspects are contradictory and it is not possible to choose one over the other; instead, they each need to be fulfilled as well as possible. In this situation ideal solutions are not always found because they are either not sought or the problems are regarded as being too complex for human capabilities. To improve this situation we propose complementing traditional design approaches with a design synthesis process based on evolutionary many-criteria optimization methods that can fulfill formalizable design requirements. In addition we show how self-organizing maps can be used to visualize many-dimensional solution spaces in an easily analyzable and comprehensible form. The system is presented using an urban planning scenario for the placement of building volumes.

show abstract

Section: Clustering Of Design Variantsmentioning

confidence: 99%

Urban Design Synthesis for Building Layouts Based on Evolutionary Many-Criteria Optimization

König

2015

International Journal of Architectural Computing

View full text Add to dashboard Cite

show abstract

“…The Contextual SOM (CSOM) supplements the standard SOM with the addition of a contextual label on the individual nodes of the resulting SOM. Prior work showed that the added contextual information allows an investigator to identify characteristics of the data set that are otherwise extremely difficult to find [15].…”

Section: Introductionmentioning

confidence: 99%

Extending parallelization of the self-organizing map by combining data and network partitioned methods

Richardson

Winer

2015

Advances in Engineering Software

Self Cite

View full text Add to dashboard Cite

Information visualization for selection in Design by Shopping

2016

View full text Add to dashboard Cite

Abstract. In Design by Shopping, designers explore the design space to gain an insight into trades and feasible and impractical solutions, as well as to learn about alternatives before optimization and selection. The design space consists of multidimensional sets of data and, in order to select the best design from amongst numerous alternatives, designers may use several different graphs. In this study, we test to find the most appropriate graph to indicate the best solution corresponding to a set of objectives represented by a design scenario (1). A further constraint is that this must be done in the shortest possible time (2). Three graph types are tested in three different design scenarios where one car has to be chosen from a total of 40. A response quality index is proposed which computes the quality of a designer's choice for any given scenario. In total, 90 tests with 30 participants were performed. The Parallel Coordinates Plot proved to be the best graph for selection in Design by Shopping.

show abstract

Visual Design Space Exploration using Contextual Self-Organizing Maps

Cited by 6 publications

References 4 publications

Urban Design Synthesis for Building Layouts Based on Evolutionary Many-Criteria Optimization

Urban Design Synthesis for Building Layouts Based on Evolutionary Many-Criteria Optimization

Extending parallelization of the self-organizing map by combining data and network partitioned methods

Information visualization for selection in Design by Shopping

Contact Info

Product

Resources

About