Abstract
Global sensitivity analysis aims to quantify the importance of model input variables for a model response. We highlight the role sensitivity analysis can play in interpretable machine learning and provide a short survey on sensitivity analysis with a focus on global variance-based sensitivity measures like Sobol’ indices and Shapley values. We discuss the Monte Carlo estimation of various Sobol’ indices as well as their graphical presentation in the so-called FANOVA graphs. Global sensitivity analysis is applied to an analytical example, a Kriging model of a piston simulator and a neural net model of the resistance of yacht hulls.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
References
Amberti, D.: mistat: Data Sets, Functions and Examples from the Book: “Modern Industrial Statistics” by Kenett, Zacks and Amberti (2018). https://CRAN.R-project.org/package=mistat. R package version 1.0-5
Boser, B.E., Guyon, I.M., Vapnik, V.N.: A training algorithm for optimal margin classifiers. In: Haussler, D. (ed.) Proceedings of the Fifth Annual Workshop on Computational Learning Theory – COLT ’92, pp. 144–152. ACM Press, New York (1992). https://doi.org/10.1145/130385.130401
Cheng, K., Lu, Z., Zhou, Y., Shi, Y., Wei, Y.: Global sensitivity analysis using support vector regression. Appl. Math. Model. 49(4), 587–598 (2017). https://doi.org/10.1016/j.apm.2017.05.026
Cortez, P., Embrechts, M.J.: Using sensitivity analysis and visualization techniques to open black box data mining models. Inf. Sci. 225(1), 1–17 (2013). https://doi.org/10.1016/j.ins.2012.10.039
Cukier, R., Levine, H., Shuler, K.: Nonlinear sensitivity analysis of multiparameter model systems. J. Comput. Phys. 26(1), 1–42 (1978). https://doi.org/10.1016/0021-9991(78)90097-9
Efron, B., Stein, C.: The jackknife estimate of variance. Ann. Stat. 9(3) (1981). https://doi.org/10.1214/AOS/1176345462
Friedman, J.H., Popescu, B.E.: Predictive learning via rule ensembles. Ann. Appl. Stat. 2(3), 916–954 (2008). https://doi.org/10.1214/07-AOAS148
Fruth, J., Roustant, O., Muehlenstaedt, T.: The fanovaGraph Package: Visualization of Interaction Structures and Construction of Block-additive Kriging Models (2013). https://hal.archives-ouvertes.fr/hal-00795229
Fruth, J., Roustant, O., Kuhnt, S.: Total interaction index: a variance-based sensitivity index for second-order interaction screening. J. Stat. Plann. Infer. 147, 212–223 (2014). https://doi.org/10.1016/j.jspi.2013.11.007
Fruth, J., Muehlenstaedt, T., Roustant, O., Jastrow, M., Kuhnt, S.: fanovaGraph: building Kriging Models from FANOVA Graphs (2020). https://CRAN.R-project.org/package=fanovaGraph. R package version 1.5
Gerritsma, J., Onnink, R., Versluis, A.: Geometry, resistance and stability of the delft systematic yacht hull series. Int. Shipbuild. Prog. 28(328), 276–297 (1981). https://doi.org/10.3233/ISP-1981-2832801
Hastie, T., Tibshirani, R.J.: Generalised additive models. In: Monographs on Statistics and Applied Probability, vol. 43. Chapman and Hall, London (1990)
Homma, T., Saltelli, A.: Importance measures in global sensitivity analysis of nonlinear models. Reliab. Eng. Syst. Saf. 52(1), 1–17 (1996). https://doi.org/10.1016/0951-8320(96)00002-6
Iooss, B., Lemaître, P.: A review on global sensitivity analysis methods. In: Dellino, G., Meloni, C. (eds.) Uncertainty Management in Simulation-Optimization of Complex Systems, Operations Research/Computer Science Interfaces Series, vol. 59, pp. 101–122. Springer US, Boston (2015). https://doi.org/10.1007/978-1-4899-7547-8_5
Iooss, B., Veiga, S.D., Janon, A., Pujol, G., with contributions from Baptiste Broto, Boumhaout, K., Delage, T., Amri, R.E., Fruth, J., Gilquin, L., Guillaume, J., Le Gratiet, L., Lemaitre, P., Marrel, A., Meynaoui, A., Nelson, B.L., Monari, F., Oomen, R., Rakovec, O., Ramos, B., Roustant, O., Song, E., Staum, J., Sueur, R., Touati, T., Weber, F.: Sensitivity: global Sensitivity Analysis of Model Outputs (2020). https://CRAN.R-project.org/package=sensitivity. R package version 1.22.1
Ishigami, T., Homma, T.: An importance quantification technique in uncertainty analysis for computer models. In: 1990 Proceedings of First International Symposium on Uncertainty Modeling and Analysis, pp. 398–403. IEEE Computer Society Press, Washington (1990). https://doi.org/10.1109/ISUMA.1990.151285
Ivanov, M., Kuhnt, S.: A parallel optimization algorithm based on FANOVA decomposition. Qual. Reliab. Eng. Int. 30(7), 961–974 (2014). https://doi.org/10.1002/qre.1710
Jansen, M.J.: Analysis of variance designs for model output. Comput. Phys. Commun. 117(1–2), 35–43 (1999). https://doi.org/10.1016/S0010-4655(98)00154-4
Kenett, R., Zacks, S., Amberti, D.: Modern Industrial Statistics: With Applications in R, MINITAB and JMP, 2nd edn. Statistics in Practice. Wiley, Chichester (2014)
Kucherenko, S., Rodriguez-Fernandez, M., Pantelides, C., Shah, N.: Monte Carlo evaluation of derivative-based global sensitivity measures. Reliab. Eng. Syst. Saf. 94(7), 1135–1148 (2009). https://doi.org/10.1016/j.ress.2008.05.006
Liu, R., Owen, A.B.: Estimating mean dimensionality of analysis of variance decompositions. J. Am. Stat. Assoc. 101(474), 712–721 (2006). https://doi.org/10.1198/016214505000001410
Lopez Gonzalez, R.: Neural networks for variational problems in engineering. PhD Thesis. Technical University of Catalonia (2008)
McCullagh, P., Nelder, J.A.: Generalized linear models. Monographs on Statistics and Applied Probability, vol. 37, 2nd edn. Chapman and Hall, London (1989)
Molnar, C.: Interpretable Machine Learning. lulu.com (2020)
Molnar, C., Casalicchio, G., Bischl, B.: Quantifying model complexity via functional decomposition for better post-hoc interpretability. In: Cellier, P., Driessens, K. (eds.) Machine Learning and Knowledge Discovery in Databases. Communications in Computer and Information Science, vol. 1167, pp. 193–204. Springer International Publishing, Cham (2020). https://doi.org/10.1007/978-3-030-43823-4_17
Muehlenstaedt, T., Roustant, O., Carraro, L., Kuhnt, S.: Data-driven kriging models based on Fanova-decomposition. Stat. Comput. 22(3), 723–738 (2012). https://doi.org/10.1007/s11222-011-9259-7
Owen, A.B.: Sobol’ indices and Shapley value. SIAM/ASA J. Uncertain. Quantif. 2(1), 245–251 (2014). https://doi.org/10.1137/130936233
Roustant, O., Fruth, J., Iooss, B., Kuhnt, S.: Crossed-derivative based sensitivity measures for interaction screening. Math. Comput. Simul. 105, 105–118 (2014). https://doi.org/10.1016/j.matcom.2014.05.005
Saltelli, A., Ratto, M., Andres, T., Campolongo, F., Cariboni, J., Gatelli, D., Saisana, M., Tarantola, S.: Global Sensitivity Analysis. The Primer. John Wiley & Sons, Ltd, Chichester (2007). https://doi.org/10.1002/9780470725184
Shapley, L.S.: A value for n-person games. In: Kuhn, H.W., Tucker, A.W. (eds.) Contributions to the Theory of Games (AM-28), vol. II, pp. 307–318. Princeton University Press, Princeton (1953). https://doi.org/10.1515/9781400881970-018
Sobol, I.M.: Sensitivity analysis for non-linear mathematical models. Math. Modeling Comput. Experiment 1(4), 407–414 (1993)
Sobol, I., Gershman, A.: On an alternative global sensitivity estimators. In: Proceedings of SAMO 1995, Belgirate, pp. 40–42 (1995)
Song, E., Nelson, B.L., Staum, J.: Shapley effects for global sensitivity analysis: theory and computation. SIAM/ASA J. Uncertain. Quantif. 4(1), 1060–1083 (2016). https://doi.org/10.1137/15M1048070
Štrumbelj, E., Kononenko, I.: Explaining prediction models and individual predictions with feature contributions. Knowl. Inf. Syst. 41(3), 647–665 (2014). https://doi.org/10.1007/s10115-013-0679-x
Venables, W.N., Ripley, B.D.: Modern Applied Statistics with S, 4th edn. Springer, New York (2002). ISBN 0-387-95457-0. http://www.stats.ox.ac.uk/pub/MASS4
Winter, E.: The Shapley value. In: Handbook of Game Theory with Economic Applications, vol. 3, pp. 2025–2054. Elsevier, Amsterdam (2002). https://doi.org/10.1016/S1574-0005(02)03016-3
Zhang, P.: A novel feature selection method based on global sensitivity analysis with application in machine learning-based prediction model. Appl. Soft Comput. 85, 105859 (2019). https://doi.org/10.1016/j.asoc.2019.105859
Acknowledgements
The financial support of the Deutsche Forschungsgemeinschaft (SFB 823, project B1) is gratefully acknowledged.
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2022 The Author(s), under exclusive license to Springer Nature Switzerland AG
About this chapter
Cite this chapter
Kuhnt, S., Kalka, A. (2022). Global Sensitivity Analysis for the Interpretation of Machine Learning Algorithms. In: Steland, A., Tsui, KL. (eds) Artificial Intelligence, Big Data and Data Science in Statistics. Springer, Cham. https://doi.org/10.1007/978-3-031-07155-3_6
Download citation
DOI: https://doi.org/10.1007/978-3-031-07155-3_6
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-031-07154-6
Online ISBN: 978-3-031-07155-3
eBook Packages: Mathematics and StatisticsMathematics and Statistics (R0)