Skip to main content

2015 | OriginalPaper | Buchkapitel

19. Interpretation of Black-Box Predictive Models

verfasst von : Vladimir Cherkassky, Sauptik Dhar

Erschienen in: Measures of Complexity

Verlag: Springer International Publishing

Aktivieren Sie unsere intelligente Suche, um passende Fachinhalte oder Patente zu finden.

search-config
loading …

Abstract

Many machine learning applications involve predictive data-analytic modeling using black-box techniques. A common problem in such studies is understanding/interpretation of estimated nonlinear high-dimensional models. Whereas human users naturally favor simple interpretable models, such models may not be practically feasible with modern adaptive methods such as Support Vector Machines (SVMs), Multilayer Perceptron Networks (MLPs), AdaBoost, etc. This chapter provides a brief survey of the current techniques for visualization and interpretation of SVM-based classification models, and then highlights potential problems with such methods. We argue that, under the VC-theoretical framework, model interpretation cannot be achieved via technical analysis of predictive data-analytic models. That is, any meaningful interpretation should incorporate application domain knowledge outside data analysis. We also describe a simple graphical technique for visualization of SVM classification models.

Sie haben noch keine Lizenz? Dann Informieren Sie sich jetzt über unsere Produkte:

Springer Professional "Wirtschaft+Technik"

Online-Abonnement

Mit Springer Professional "Wirtschaft+Technik" erhalten Sie Zugriff auf:

  • über 102.000 Bücher
  • über 537 Zeitschriften

aus folgenden Fachgebieten:

  • Automobil + Motoren
  • Bauwesen + Immobilien
  • Business IT + Informatik
  • Elektrotechnik + Elektronik
  • Energie + Nachhaltigkeit
  • Finance + Banking
  • Management + Führung
  • Marketing + Vertrieb
  • Maschinenbau + Werkstoffe
  • Versicherung + Risiko

Jetzt Wissensvorsprung sichern!

Springer Professional "Technik"

Online-Abonnement

Mit Springer Professional "Technik" erhalten Sie Zugriff auf:

  • über 67.000 Bücher
  • über 390 Zeitschriften

aus folgenden Fachgebieten:

  • Automobil + Motoren
  • Bauwesen + Immobilien
  • Business IT + Informatik
  • Elektrotechnik + Elektronik
  • Energie + Nachhaltigkeit
  • Maschinenbau + Werkstoffe




 

Jetzt Wissensvorsprung sichern!

Springer Professional "Wirtschaft"

Online-Abonnement

Mit Springer Professional "Wirtschaft" erhalten Sie Zugriff auf:

  • über 67.000 Bücher
  • über 340 Zeitschriften

aus folgenden Fachgebieten:

  • Bauwesen + Immobilien
  • Business IT + Informatik
  • Finance + Banking
  • Management + Führung
  • Marketing + Vertrieb
  • Versicherung + Risiko




Jetzt Wissensvorsprung sichern!

Literatur
1.
Zurück zum Zitat Adams, N.M., Hand, D.J.: Improving the practice of classifier performance assessment. Neural Comput. 12(2), 305–311 (2000)CrossRef Adams, N.M., Hand, D.J.: Improving the practice of classifier performance assessment. Neural Comput. 12(2), 305–311 (2000)CrossRef
3.
Zurück zum Zitat Barakat, N., Bradley, A.: Rule-extraction from support vector machines: a review. Neurocomputing 74(1–3), 178–190 (2010)CrossRef Barakat, N., Bradley, A.: Rule-extraction from support vector machines: a review. Neurocomputing 74(1–3), 178–190 (2010)CrossRef
4.
Zurück zum Zitat Boyd, S., Vandenberghe, L.: Convex Optimization. Cambridge University Press, New York (2004)MATHCrossRef Boyd, S., Vandenberghe, L.: Convex Optimization. Cambridge University Press, New York (2004)MATHCrossRef
5.
Zurück zum Zitat Bradley, A.: The use of the area under the ROC curve in the evaluation of machine learning algorithms. Pattern Recognit. 30(7), 1145–1159 (1997)CrossRef Bradley, A.: The use of the area under the ROC curve in the evaluation of machine learning algorithms. Pattern Recognit. 30(7), 1145–1159 (1997)CrossRef
7.
Zurück zum Zitat Caragea, D., Cook, D., Honavar, V.G.: Gaining insights into support vector machine pattern classifiers using projection-based tour methods. In: Proceedings of KDD, pp. 251–256 (2001) Caragea, D., Cook, D., Honavar, V.G.: Gaining insights into support vector machine pattern classifiers using projection-based tour methods. In: Proceedings of KDD, pp. 251–256 (2001)
8.
Zurück zum Zitat Cherkassky, V.: Predictive learning, knowledge discovery and philosophy of science (invited lecture). In: Lin, J., et al. (eds.) Advances in Computational Intelligence, vol. 7311, pp. 209–233. Springer, Berlin (2012)CrossRef Cherkassky, V.: Predictive learning, knowledge discovery and philosophy of science (invited lecture). In: Lin, J., et al. (eds.) Advances in Computational Intelligence, vol. 7311, pp. 209–233. Springer, Berlin (2012)CrossRef
10.
Zurück zum Zitat Cherkassky, V., Dhar, S.: Simple method for interpretation of high-dimensional nonlinear SVM classification models. In: Proceedings of the 2010 International Conference on Data Mining (DMIN 2010), pp. 267–272 (2010) Cherkassky, V., Dhar, S.: Simple method for interpretation of high-dimensional nonlinear SVM classification models. In: Proceedings of the 2010 International Conference on Data Mining (DMIN 2010), pp. 267–272 (2010)
11.
Zurück zum Zitat Cherkassky, V., Dhar, S.: Market timing of international mutual funds: a decade after the scandal. In: Proceedings of Computational Intelligence for Financial Engineering and Economics, pp. 1–8 (2012) Cherkassky, V., Dhar, S.: Market timing of international mutual funds: a decade after the scandal. In: Proceedings of Computational Intelligence for Financial Engineering and Economics, pp. 1–8 (2012)
12.
Zurück zum Zitat Cherkassky, V., Dhar, S., Dai, W.: Practical conditions for effectiveness of the universum learning. IEEE Trans. Neural Netw. 22(8), 1241–1255 (2011)CrossRef Cherkassky, V., Dhar, S., Dai, W.: Practical conditions for effectiveness of the universum learning. IEEE Trans. Neural Netw. 22(8), 1241–1255 (2011)CrossRef
13.
Zurück zum Zitat Cherkassky, V., Mulier, F.: Learning from Data: Concepts, Theory, and Methods. Wiley, New York (1998)MATH Cherkassky, V., Mulier, F.: Learning from Data: Concepts, Theory, and Methods. Wiley, New York (1998)MATH
14.
Zurück zum Zitat Cherkassky, V., Mulier, F.: Learning from Data: Concepts, Theory, and Methods, 2nd edn. Wiley, New York (2007)CrossRef Cherkassky, V., Mulier, F.: Learning from Data: Concepts, Theory, and Methods, 2nd edn. Wiley, New York (2007)CrossRef
15.
Zurück zum Zitat Cook, D., Swayne, D.F.: Interactive and Dynamic Graphics for Data Analysis: With Examples Using R and GGobi. Springer, New York (2007)CrossRef Cook, D., Swayne, D.F.: Interactive and Dynamic Graphics for Data Analysis: With Examples Using R and GGobi. Springer, New York (2007)CrossRef
16.
17.
Zurück zum Zitat Fisher, R.: The logic of inductive inference. J. R. Stat. Soc. 98(1), 39–82 (1935)CrossRef Fisher, R.: The logic of inductive inference. J. R. Stat. Soc. 98(1), 39–82 (1935)CrossRef
18.
Zurück zum Zitat Fisher, R.: The use of multiple measurements in taxonomic problems. Ann. Eugen. 7(2), 179–188 (1935)CrossRef Fisher, R.: The use of multiple measurements in taxonomic problems. Ann. Eugen. 7(2), 179–188 (1935)CrossRef
19.
Zurück zum Zitat Guyon, I., Weston, J., Barnhill, S., Vapnik, V.: Gene selection for cancer classification using support vector machines. Mach. Learn. 46(1–3), 389–422 (2002)MATHCrossRef Guyon, I., Weston, J., Barnhill, S., Vapnik, V.: Gene selection for cancer classification using support vector machines. Mach. Learn. 46(1–3), 389–422 (2002)MATHCrossRef
21.
Zurück zum Zitat Hand, D.J., Mannila, H., Smyth, P.: Principles of Data Mining. MIT Press, Cambridge (2001) Hand, D.J., Mannila, H., Smyth, P.: Principles of Data Mining. MIT Press, Cambridge (2001)
22.
Zurück zum Zitat Hastie, T., Tibshirani, R., Friedman, J.: The Elements of Statistical Learning: Data Mining, Inference and Prediction. Springer, New York (2001)CrossRef Hastie, T., Tibshirani, R., Friedman, J.: The Elements of Statistical Learning: Data Mining, Inference and Prediction. Springer, New York (2001)CrossRef
24.
Zurück zum Zitat Martens, D., Baesens, B., Gestel, T.: Decompositional rule extraction from support vector machines by active learning. IEEE Trans. Knowl. Data Eng. 21(2), 178–191 (2009)CrossRef Martens, D., Baesens, B., Gestel, T.: Decompositional rule extraction from support vector machines by active learning. IEEE Trans. Knowl. Data Eng. 21(2), 178–191 (2009)CrossRef
25.
Zurück zum Zitat Platt, J.C.: Probabilistic outputs for support vector machines and comparisons to regularized likelihood methods. In: Smola, A., et al. (eds.) Advances in Large Margin Classifiers, pp. 61–74. MIT Press, Cambridge (2000) Platt, J.C.: Probabilistic outputs for support vector machines and comparisons to regularized likelihood methods. In: Smola, A., et al. (eds.) Advances in Large Margin Classifiers, pp. 61–74. MIT Press, Cambridge (2000)
26.
Zurück zum Zitat Poulet, F.: SVM and graphical algorithms: a cooperative approach. In: Proceedings of the Fourth IEEE International Conference on Data Mining, pp. 499–502 (2004) Poulet, F.: SVM and graphical algorithms: a cooperative approach. In: Proceedings of the Fourth IEEE International Conference on Data Mining, pp. 499–502 (2004)
28.
Zurück zum Zitat Suykens, J.A.K., Van Gestel, T., de Brabanter, J., de Moor, B., Vandewalle, J.: Least Squares Support Vector Machines. World Scientific, Singapore (2002)MATHCrossRef Suykens, J.A.K., Van Gestel, T., de Brabanter, J., de Moor, B., Vandewalle, J.: Least Squares Support Vector Machines. World Scientific, Singapore (2002)MATHCrossRef
29.
Zurück zum Zitat Tan, P.N., Steinbach, M., Kumar, V.: Introduction to Data Mining. Addison-Wesley, Boston (2006) Tan, P.N., Steinbach, M., Kumar, V.: Introduction to Data Mining. Addison-Wesley, Boston (2006)
31.
32.
Zurück zum Zitat Vapnik, V.N.: : Estimation of Dependences Based on Empirical Data. Empirical Inference Science: Afterword of 2006. Springer, New York (2006) Vapnik, V.N.: : Estimation of Dependences Based on Empirical Data. Empirical Inference Science: Afterword of 2006. Springer, New York (2006)
33.
Zurück zum Zitat Wang, X., Wu, S., Li, Q.: SVMV—a novel algorithm for the visualization of SVM classification results. In: Wang, J., et al. (eds.) Advances in Neural Networks. Lecture Notes in Computer Science, vol. 3971, pp. 968–973. Springer, Berlin (2006) Wang, X., Wu, S., Li, Q.: SVMV—a novel algorithm for the visualization of SVM classification results. In: Wang, J., et al. (eds.) Advances in Neural Networks. Lecture Notes in Computer Science, vol. 3971, pp. 968–973. Springer, Berlin (2006)
34.
Zurück zum Zitat Witten, I.H., Frank, E.: Data Mining: Practical Machine Learning Tools and Techniques. Elsevier, Amsterdam (2005) Witten, I.H., Frank, E.: Data Mining: Practical Machine Learning Tools and Techniques. Elsevier, Amsterdam (2005)
35.
Zurück zum Zitat Zitzewitz, E.: Who cares about shareholders? Arbitrage proofing mutual funds. J. Law Econ. Organ. 19(2), 245–280 (2003)CrossRef Zitzewitz, E.: Who cares about shareholders? Arbitrage proofing mutual funds. J. Law Econ. Organ. 19(2), 245–280 (2003)CrossRef
Metadaten
Titel
Interpretation of Black-Box Predictive Models
verfasst von
Vladimir Cherkassky
Sauptik Dhar
Copyright-Jahr
2015
DOI
https://doi.org/10.1007/978-3-319-21852-6_19