Related Experiment Videos
Software systems to control sensitivity studies of Monte Carlo simulation models
1Division of Health Computer Sciences, University of Minnesota, Minneapolis 55455.
Computers and Biomedical Research, an International Journal
|December 1, 1988
Summary
This study introduces a software system to automate repetitive simulations for Monte Carlo models, enhancing the efficiency and plausibility of sensitivity analysis. The system improves modularity and reduces human intervention in complex computational studies.
Area of Science:
- Computational science
- Epidemiology
- Software engineering
Background:
- Sensitivity analysis of Monte Carlo models requires repetitive simulations.
- Analyzing variations in simulated outcomes is complex and time-consuming.
- Current methods often involve significant human intervention.
Purpose of the Study:
- To design a software system for controlling interacting programs in sensitivity analysis.
- To reduce human intervention in repetitive simulation tasks.
- To improve the modularity and efficiency of sensitivity studies for Monte Carlo models.
Main Methods:
- Developed a software system to manage and execute component packages.
- Implemented interprogram control and communication mechanisms for independent program operation.
- Isolated system-wide assumptions within a dedicated program.
Main Results:
- The software system effectively controls execution of interacting programs.
- Component programs can operate independently of data sources and execution environments.
- Demonstrated applicability to sensitivity analysis of Monte Carlo epidemic models.
Conclusions:
- The methodology enhances modularity at the executable program level.
- The system increases the plausibility and efficiency of sensitivity studies.
- Automated control reduces the need for human intervention in complex simulations.