Accurate analysis of large-scale stochastic simulations is very time consuming. The paper addresses computational issues in fitting and generating error measures of simulation metamodels and demonstrates the merit of high-performance computing in Python.
Computationally and Statistically Efficient Model Fitting Techniques
Download Resources
PDF Accessibility
One or more of the PDF files on this page fall under E202.2 Legacy Exceptions and may not be completely accessible. You may request an accessible version of a PDF using the form on the Contact Us page.
In large-scale stochastic simulations, analysis with sufficient accuracy is often extremely time consuming. The complexity of the analysis is exacerbated with increasing dimensionality of the parameter space and sudden abruptness in the topology of the input-output response surface. This paper addresses computational issues in fitting and generating error measures of simulation metamodels and the merit of high-performance computing in Python is demonstrated. A systematic comparison (of speed) is made implementing different programming languages including MATLAB, R and Python as well as using different computing architectures including high-performing laptops and high-power parallel processing clusters. The experimentation is discussed in this paper using a simple scenario, and activities are being pursued to study other scenarios with varying complexities that will be reported at the conference.