Jump to content

Draft:Global Sensitivity Analysis. The Primer

From Wikipedia, the free encyclopedia
Global Sensitivity Analysis. The Primer
AuthorsAndrea Saltelli, Marco Ratto, Terry Andres, Francesca Campolongo, Jessica Cariboni, Debora Gatelli, Michaela Saisana, Stefano Tarantola
LanguageEnglish
SubjectsMathematical modelling
Applied Statistics
Model validationImpact assessment
Evidence Based Policy
Publisher John Wiley & Sons
Publication date
18 December 2007
Pages304
ISBN978-0-470-05997-5

"Global Sensitivity Analysis. The Primer"[1] by Andrea Saltelli and others is an introduction to sensitivity analysis of model output, a discipline that studies how the uncertainty in model input and model assumptions propagates to model output and model-based inference. The volume was published in December 2007 by John Wiley & Sons. The same publisher offered a Chinese translation in 2018. [2]

Synopsis

[edit]

The volume offers an introduction to the validation of mathematical models, trying to answer questions such as[3] ‘’How sensitive are results to an assumed input value?’’; ‘’What variables are driving conclusions?’’; ‘’Can I simplify this model?’’; ‘’What parameter levels will lead to a desired outcome?’’ Exercises and solutions are provided at the end of each chapter. [3]



Table of content[4]
Chapter Title Content
1. Introduction to sensitivity analysis A philosophical introduction to models. How to read the book.
2. Experimental design Can methods of statistical Design of experiments be applied to mathematical modelling?
3. Elementary effects method The Morris method and its variants
4. Variance-based methods Methods based on decomposing the variance of the output
5. Factor mapping and metamodelling (with Peter Young) An introduction to metamodeling and Monte Carlo filtering
6. Sensitivity analysis: From theory to practice More applications of sensitivity analysis with policy example

Reception

[edit]

According to[3] the book is to be praised for clarity of exposition, wealth of examples, and solved exercises, while a limitation is that it appears written more for mathematical modelling and simulation than for statistical modelling and does not cover ‘sensitivity of, for example, survival estimates to the assumption of noninformative censoring or odds ratios to the assumption of no unmeasured confounders’, though another reviewer[5] finds the book instructive for statistical models as well. A quote of the book mentioned in the review[3] is:

If modelling is a craft and models cannot be proven true, then the modeller has a moral obligation, and indeed it is in the modeller’s own practical interest, to be as rigorous as possible when assessing the robustness of model inference.

For[4] the volume takes a practical approach with motivation for sensitivity analysis, reviews required statistical concepts, and provides a guide to potential applications in several chapters and more diffusely in chapter 6. A limit of the book is that in focusing on global sensitivity analysis methods it does not treat local methods for sensitivity analysis such as Samprit Chatterjee’s ‘Sensitivity Analysis in Linear Regression’[6], however, it is a welcome addition to its sister volume Sensitivity Analysis in Practice: A Guide to Assessing Scientific Models.[7][8][9]

For[5]

The type of sensitivity analysis that the authors speak of is related to models in general and statistical models in particular. How should the assumptions of the model be tested? Are there computer or experimental designs that can be useful in determining how sensitive the model is to departure from the assumptions?

The philosophical introduction about the nature of models useful for a statistical readership. [5] A statistician might be particularly interested in the Morris_method of chapter 3 and in the variance-based methods of Ylia M. Sobol’ (to whom the book is dedicated) in chapter 4. [5]

References

[edit]
  1. ^ Saltelli, A., Ratto, M., Andres, T. H., Campolongo, F., Cariboni, J., Gatelli, D., Saisana, M., Tarantola, S. (2008). Global sensitivity analysis : the primer. John Wiley. ISBN 0-470-05997-4.
  2. ^ Wu, Q., Ding, Y., Yi, M., Fan, Q. (2018). Global sensitivity analysis( Chinese version). Tsinghua University Publisher. ISBN 9787302485551.
  3. ^ a b c d Shepherd, B. E. (1 December 2009). "Global Sensitivity Analysis. The Primer by SALTELLI, A., RATTO, M., ANDRES, T., CAMPOLONGO, F., CARIBONI, J., GATELLI, D., SAISANA, M., and TARANTOLA, S.". Biometrics. 65 (4): 1311–1312. doi:10.1111/j.1541-0420.2009.01343_7.x. ISSN 0006-341X.
  4. ^ a b Liu, S. (2008). "Global Sensitivity Analysis: The Primer by Andrea Saltelli, Marco Ratto, Terry Andres, Francesca Campolongo, Jessica Cariboni, Debora Gatelli, Michaela Saisana, Stefano Tarantola". International Statistical Review. 76 (3). International Statistical Institute: 452–452. ISSN 0306-7734.
  5. ^ a b c d Chernick, M. (November 2008). "Global Sensitivity Analysis, the Primer". Technometrics. 50 (4). American Society for Quality: 548. ISSN 0040-1706..
  6. ^ Chatterjee, Samprit; Hadi, Ali S. (2009). Sensitivity Analysis in Linear Regression. John Wiley & Sons. pp. 54–59. ISBN 9780470317426.
  7. ^ Saltelli, A., Tarantola, S., Campolongo, F., Ratto, M. (February 2004). Sensitivity Analysis in Practice. John Wiley & Sons, Ltd. doi:10.1002/0470870958. ISBN 0-470-87093-1.
  8. ^ Paruggia, M. (2006). "Sensitivity Analysis in Practice: A Guide to Assessing Scientific Models". Journal of the American Statistical Association. 101 (473): 398–399. doi:10.1198/jasa.2006.s80.
  9. ^ McCulloch, A. (1 March 2005). "Sensitivity Analysis in Practice: A Guide to Assessing Scientific Models". Journal of the Royal Statistical Society Series A: Statistics in Society. 168 (2): 466. doi:10.1111/j.1467-985X.2005.358_16.x. ISSN 0964-1998.
[edit]

See also

[edit]
[edit]