方法对比

	确定性敏感性分析	蒙特卡洛模拟	Stochastic Sensitivity Analysis
领域	仿真	决策	仿真
方法族	Process / pipeline	MCDM	Process / pipeline
起源年份	1950s–1970s (formalized)	1949	1990s–2000s
提出者	Saltelli, A. et al.; widely formalized across operations research and health economics	Metropolis, N., Ulam, S.	Saltelli, A. et al.; Claxton, K. et al. (health economics stream)
类型	Parameter variation / robustness testing	Robustness wrapper — Monte Carlo uncertainty propagation	Probabilistic uncertainty quantification technique
开创性文献	Saltelli, A., Tarantola, S., Campolongo, F., & Ratto, M. (2004). Sensitivity Analysis in Practice: A Guide to Assessing Scientific Models. John Wiley & Sons, Chichester. ISBN: 9780470870938	Metropolis, N., Ulam, S. (1949). The Monte Carlo method. Journal of the American Statistical Association	Saltelli, A., Ratto, M., Andres, T., Campolongo, F., Cariboni, J., Gatelli, D., Saisana, M., Tarantola, S. (2008). Global Sensitivity Analysis: The Primer. Wiley. ISBN: 9780470059975
别名	DSA, One-Way Sensitivity Analysis, Tornado Diagram Analysis, Parametric Sensitivity Analysis	—	PSA, Probabilistic Sensitivity Analysis, Stochastic SA, Monte Carlo Sensitivity Analysis
相关	2	0	5
摘要	Deterministic Sensitivity Analysis (DSA) tests how model outputs change when individual or combined input parameters are varied across plausible ranges, one at a time or in structured combinations, without invoking probabilistic sampling. It is the standard approach in economic modeling, decision trees, and mathematical programming to identify which parameters drive conclusions and to demonstrate model robustness to regulators, reviewers, and stakeholders.	MONTE-CARLO-SIMULATION (Monte Carlo Simulation — Stochastic uncertainty propagation through MCDM model) is a ranking multi-criteria decision-making (MCDM) method introduced by Metropolis, N., Ulam, S. in 1949. It turns a decision matrix of alternatives scored on multiple criteria into a structured, reproducible result.	Stochastic Sensitivity Analysis (PSA) extends classical one-at-a-time sensitivity testing by representing uncertain model inputs as probability distributions and propagating them through the model via Monte Carlo sampling. The result is a full distribution of possible outputs, together with rankings of which inputs drive output variance the most — enabling robust, evidence-grounded conclusions under uncertainty.