...
  • 主题
高级搜索  >
历史搜索 清空历史
首页> 外文期刊>Journal of Universal Computer Science >The Role of Verification and Validation Techniques within Visual Analytics
【24h】

The Role of Verification and Validation Techniques within Visual Analytics

机译：验证和验证技术在Visual Analytics中的作用

获取原文
   

获取外文期刊封面封底 >>

       
页面导航
  • 摘要
  • 著录项
  • 相似文献
  • 相关主题

摘要

We suggest to widen the focus of the scientific computations community from an isolated consideration of reliable numerical algorithms using standardized arithmetic to a broad user-centered system modeling and simulation approach relying on an appropriate verification and validation (V&V) design. Most V&V works rarely consider human-related issues specifically. However, modern applications generate and employ huge amounts of heterogeneous data and usually exhibit high complexity - challenges that are best tackled by augmenting human reasoning with automated techniques. That is, novel visual and collaborative approaches are needed to interpret the results, which has to be accounted for in the general V&V procedure. This should include an assessment of (meta-) data and code/outcome quality, selection of methods to propagate and bound uncertainty and, lastly, formally rigorous validation efforts. We present an approach to reliable visual analytics (i.e., analytics subjected to this V&V assessment), which can in turn contribute to the overall V&V procedure after that. Two use cases illustrate the potential of the introduced framework for reliable visual analytics.
机译：我们建议扩大科学计算界的关注范围，从孤立地考虑使用标准化算术的可靠数值算法到依赖适当验证和验证（V＆V）设计的以用户为中心的广泛系统建模和仿真方法。大多数V＆V作品很少专门考虑与人有关的问题。但是，现代应用程序会生成并使用大量的异构数据，并且通常表现出很高的复杂性-最好通过使用自动化技术来增强人类推理来解决这些挑战。也就是说，需要新颖的视觉和协作方法来解释结果，这必须在常规V＆V程序中加以考虑。这应该包括对（元）数据和代码/结果质量的评估，选择用于传播和限制不确定性的方法，最后是正式严格的验证工作。我们提出了一种可靠的视觉分析方法（即经过此V＆V评估的分析），进而可以为之后的整个V＆V程序做出贡献。两个用例说明了引入的框架进行可靠的可视化分析的潜力。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱：kefu@zhangqiaokeyan.com

京公网安备：11010802029741号 ICP备案号：京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有

  • 客服微信

  • 服务号