Enabling uncertainty quantification (UQ) for a computational model can be a complex process, often exceeding the domain expertise of its developer or user. To address this challenge, we developed the mUQSA platform, which streamlines typical uncertainty quantification scenarios through a modern web-based environment and pre-configured automated workflows that leverage large-scale computing resources for state-of-the-art UQ techniques. mUQSA is designed with two key objectives: to support users in analyzing how input uncertainties propagate through their models and to identify the input parameters that most significantly influence outcomes through sensitivity analysis (SA). The particular strength of mUQSA lies in its accessibility: offered as a Software-as-a-Service (SaaS) solution, mUQSA eliminates the need for installation or configuration of software, enabling users to access UQ/SA methods directly from a web browser. Its user-friendly, intuitive interfaces for defining use cases and displaying results guarantee high accessibility, while a robust execution layer, supported by HPC resources, ensures mUQSA can handle even very complex computational models.

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

mUQSA – An Online Service for Uncertainty Quantification and Sensitivity Analysis

  • Bartosz Bosak,
  • Piotr Kopta,
  • Michał Kulczewski,
  • Tomasz Piontek

摘要

Enabling uncertainty quantification (UQ) for a computational model can be a complex process, often exceeding the domain expertise of its developer or user. To address this challenge, we developed the mUQSA platform, which streamlines typical uncertainty quantification scenarios through a modern web-based environment and pre-configured automated workflows that leverage large-scale computing resources for state-of-the-art UQ techniques. mUQSA is designed with two key objectives: to support users in analyzing how input uncertainties propagate through their models and to identify the input parameters that most significantly influence outcomes through sensitivity analysis (SA). The particular strength of mUQSA lies in its accessibility: offered as a Software-as-a-Service (SaaS) solution, mUQSA eliminates the need for installation or configuration of software, enabling users to access UQ/SA methods directly from a web browser. Its user-friendly, intuitive interfaces for defining use cases and displaying results guarantee high accessibility, while a robust execution layer, supported by HPC resources, ensures mUQSA can handle even very complex computational models.