ParaGlide: Interactive Parameter Space Partitioning for Computer Simulations
ParaGlide: Interactive Parameter Space Partitioning for Computer Simulations
复制标题
DOI：
10.1109/tvcg.2013.61
复制
发表时间：
2013-09-01
影响因子：
5.2
通讯作者：
Saad, Ahmed
中科院分区：
文献类型：
--
作者：
Bergner, Steven;Sedlmair, Michael;Saad, Ahmed
本文介绍了一个用于多维仿真模型参数空间交互探索的可视化系统ParaGlide。为了获得正确的参数配置，模型开发人员经常必须在设置输入参数和定性地判断模型的结果之间来回切换。然而，目前最先进的工具和实践无法提供一种系统的方法来探索这些参数空间，因此对参数配置做出明智的决定是一项繁琐且工作量大的任务。ParaGlide通过使用基于区域的用户界面进行参数采样来指导数据生成，然后将模型的输入参数空间划分为代表不同输出行为的分区，从而努力克服这一缺点。特别是，我们发现参数空间划分可以帮助模型开发人员更好地理解可能的高维模型输出之间的定性差异。此外，它还提供了参数灵敏度的信息，便于模型的比较。我们与来自三个不同领域的专家密切合作开发了ParaGlide，他们都参与了各自领域的新模型开发。我们首先分析了六位领域专家的当前实践，并得出了一组任务和设计需求，然后进行了以用户为中心的设计过程，最后进行了三个纵向深入的案例研究，强调了我们方法的有效性。
In this paper, we introduce ParaGlide, a visualization system designed for interactive exploration of parameter spaces of multidimensional simulation models. To get the right parameter configuration, model developers frequently have to go back and forth between setting input parameters and qualitatively judging the outcomes of their model. Current state-of-the-art tools and practices, however, fail to provide a systematic way of exploring these parameter spaces, making informed decisions about parameter configurations a tedious and workload-intensive task. ParaGlide endeavors to overcome this shortcoming by guiding data generation using a region-based user interface for parameter sampling and then dividing the model's input parameter space into partitions that represent distinct output behavior. In particular, we found that parameter space partitioning can help model developers to better understand qualitative differences among possibly high-dimensional model outputs. Further, it provides information on parameter sensitivity and facilitates comparison of models. We developed ParaGlide in close collaboration with experts from three different domains, who all were involved in developing new models for their domain. We first analyzed current practices of six domain experts and derived a set of tasks and design requirements, then engaged in a user-centered design process, and finally conducted three longitudinal in-depth case studies underlining the usefulness of our approach.