Skip to main content
SpringerLink
Log in
Book cover

Decision Making: Uncertainty, Imperfection, Deliberation and Scalability pp 55–89Cite as

On the Origins of Imperfection and Apparent Non-rationality

  • Chapter
  • First Online:

Part of the book series: Studies in Computational Intelligence ((SCI,volume 538))

Abstract

Decision making (DM) is a preferences-driven choice among available actions. Under uncertainty, Savage’s axiomatisation singles out Bayesian DM as the adequate normative framework. It constructs strategies generating the optimal actions, while assuming that the decision maker rationally tries to meet her preferences. Descriptive DM theories have observed numerous deviations of the real DM from normative recommendations. The explanation of decision-makers’ imperfection or non-rationality, possibly followed by rectification, is the focal point of contemporary DM research. This chapter falls into this stream and claims that the neglecting a part of the behaviour of the closed DM loop is the major cause of these deviations. It inspects DM subtasks in which this claim matters and where its consideration may practically help. It deals with: (i) the preference elicitation; (ii) the “non-rationality” caused by the difference of preferences declared and preferences followed; (iii) the choice of proximity measures in knowledge and preferences fusion; (iv) ways to a systematic design of approximate DM; and (v) the control of the deliberation effort spent on a DM task via sequential DM. The extent of the above list indicates that the discussion offers more open questions than answers, however, their consideration is the key element of this chapter. Their presentation is an important chapter’s ingredient.

This is a preview of subscription content, log in via an institution.

Chapter
USD   29.95
Price excludes VAT (USA)
  • Available as PDF
  • Read on any device
  • Instant download
  • Own it forever
eBook
USD   84.99
Price excludes VAT (USA)
  • Available as EPUB and PDF
  • Read on any device
  • Instant download
  • Own it forever
Hardcover Book
USD   109.99
Price excludes VAT (USA)
  • Durable hardcover edition
  • Dispatched in 3 to 5 business days
  • Free shipping worldwide - see info

Tax calculation will be finalised at checkout

Purchases are for personal use only

Learn about institutional subscriptions

Notes

  1. 1.

    Savage [68] calls it a small world. Alternative terms like system, plant, object are used.

  2. 2.

    A pd is the Radon-Nikodým derivative of a probabilistic, randomness-modelling measure.

  3. 3.

    The KLD has many names. Relative entropy and cross entropy [70] are the most common.

  4. 4.

    When ignorance includes non-constant internals, Bayesian learning used below becomes stochastic filtering [30]. If moreover, the decision maker’s preferences depend on an action-dependent internal state, the stochastic control problem arises [41]. This general case is not treated here as it complicates explanations without offering any conceptual shift.

  5. 5.

    Further on, the superscript \(^{\star }\) marks pds and actions arising from this ideal closed-loop model.

  6. 6.

    The quest for simple final formulas has motivated a slightly non-standard choice of the “directions” of the ordering operators \(\preceq \), \(\ge \) and \(\prec \), \(>\).

  7. 7.

    The existence of such pairs can be assumed without loss of generality. Indeed, no non-trivial decision task arises if all comparable pairs of behaviours in the original decision-maker-specified partial ordering are equivalent.

  8. 8.

    The mapping \({\mathsf {R}}_{{\mathsf {S}}}\) is common to decision makers differing only in preferences among behaviours.

  9. 9.

    The functional is local if its value on \({\mathsf {\Lambda }}\), artificially written as the sum \({\mathsf {\Lambda }}_{1}+{\mathsf {\Lambda }}_{2}\) of functions \({\mathsf {\Lambda }}_{1},\,{\mathsf {\Lambda }}_{2}\) fulfilling \({\mathsf {\Lambda }}_{1}{\mathsf {\Lambda }}_{2}=0\), is the sum of its values on \({\mathsf {\Lambda }}_{1}\) and \({\mathsf {\Lambda }}_{2}\).

  10. 10.

    The measure serves to all DM tasks facing the same uncertainty. The function \({\mathsf {U}}\) models risk awareness, neutrality or proneness. The function \({\mathsf {U}}\), \({\mathsf {C}}\)-almost surely increasing in its first argument, guarantees that the optimal strategy \({\mathsf {S}}^{o}\) selected from the considered subset of \({\pmb {{{{\mathsf {S}}}}}}\) is not dominated It means that it cannot happen that within this subset there is a strategy \({\mathsf {S}}^{d}\) such that \({\mathsf {\Lambda }}_{{\mathsf {S}}^{d}}(u)\le {\mathsf {\Lambda }}_{{\mathsf {S}}^{o}}(u)\) on \({\pmb {{{u}}}}\) with the sharp inequality on a subset of \({\pmb {{{u}}}}\) of a positive \({\mathsf {C}}\) measure.

  11. 11.

    Giarlotta and Greco [22] represents non-Bayesian set-ups dealing with sets of orderings without a quest for a unique completion.

  12. 12.

    A decision maker interacts with customers in order to influence them in a desirable direction, for instance, to buy a specific product or services. However, even the form of the questionnaire influences the customers: typically, two different ways of posing logically the same question often provide quite different answers. This quantum-mechanics-like effect should be properly modelled.

  13. 13.

    The vast majority of complex technological processes, which should be modelled by high-dimensional nonlinear stochastic partial differential equations with non-smooth boundary conditions, are controlled by proportional-integral-derivative controllers corresponding to simple linear, second order difference equations used as the environment model.

  14. 14.

    The adopted notation \(a^{\star }\) stresses that this action value serves for the construction of \({\mathsf {C}}^{\star }\).

  15. 15.

    \({\mathsf {S}}^{\star }_{0}(a_{t}|o_{t},k_{t-1})\) and \(a^{\star }_{t}(k_{t-1})\) are independent of \(o_{t}\), i.e. \({\mathsf {S}}^{\star }_{}(a_{t}|o_{t},k_{t-1})={\mathsf {S}}^{\star }(a_{t}|k_{t-1})\), see (3.21).

  16. 16.

    The condition \(z_{t}=1\) stresses that the optimisation is performed: it is not stopped.

References

  1. Barndorff-Nielsen, O.: Information and Exponential Families in Statistical Theory. Wiley, New York (1978)

    MATH  Google Scholar 

  2. Belda, K.: Probabilistically tuned LQ control for mechatronic applications (paper version). AT&P J. 9(2), 19–24 (2009)

    MathSciNet  Google Scholar 

  3. Bellman, R.: Adaptive Control Processes. Princeton University Press, Princeton (1961)

    MATH  Google Scholar 

  4. Bernardo, J.: Expected information as expected utility. Ann. Stat. 7(3), 686–690 (1979)

    Article  MATH  MathSciNet  Google Scholar 

  5. Bertsekas, D.: Dynamic Programming and Optimal Control. Athena Scientific, Belmont (2001)

    MATH  Google Scholar 

  6. Bohlin, T.: Interactive System Identification: Prospects and Pitfalls. Springer, New York (1991)

    Book  MATH  Google Scholar 

  7. Boutilier, B.: A POMDP formulation of preference elicitation problems. In: Proceedings of the 18th National Conference on AI, AAAI-2002, pp. 239–246. Edmonton (2002)

    Google Scholar 

  8. Boutilier, C., Drummond, J., Lu, T.: Preference elicitation for social choice: a study in stable matching and voting. In: Guy, T., Kárný, M. (eds.) Proceedings of the 3rd International Workshop on Scalable Decision Making, ECML/PKDD 2013, ÚTIA AVČR, Prague (2013)

    Google Scholar 

  9. Campenhout, J.V., Cover, T.: Maximum entropy and conditional probability. IEEE Trans. Inf. Theory 27(4), 483–489 (1981)

    Article  MATH  Google Scholar 

  10. Cappe, O., Godsill, S., Moulines, E.: An overview of existing methods and recent advances in sequential Monte Carlo. Proc. IEEE 95(5), 899–924 (2007). doi:10.1109/JPROC.2007.893250

    Article  Google Scholar 

  11. Chen, L., Pu, P.: Survey of preference elicitation methods. Technical Report IC/2004/67, Human Computer Interaction Group Ecole Politechnique Federale de Lausanne (EPFL), CH-1015 Lausanne (2004)

    Google Scholar 

  12. Conlisk, J.: Why bounded rationality? J. Econ. Behav. Organ. 34(2), 669–700 (1996)

    Google Scholar 

  13. Debreu, G.: Representation of a preference ordering by a numerical function. In: Thrall, R., Coombs, C., Davis, R. (eds.) Decision Processes. Wiley, New York (1954)

    Google Scholar 

  14. DeWitt, B., Graham, N.: The Many-Worlds Interpretation of Quantum Mechanics. Princeton University Press, Princeton (1973)

    Google Scholar 

  15. Dvurečenskij, A.: Gleasons Theorem and Its Applications, Mathematics and Its Applications, vol. 60. Kluwer, Bratislava (1993)

    Book  Google Scholar 

  16. Efron, B.: Why isn’t everyone a Bayesian. Am. Stat. 40(1), 1–11 (1986)

    MATH  MathSciNet  Google Scholar 

  17. Feldbaum, A.: Theory of dual control. Autom. Remote Control 21(9), 874–880 (1960)

    MathSciNet  Google Scholar 

  18. Ferguson, T.: A Bayesian analysis of some nonparametric problems. Ann. Stat. 1, 209–230 (1973)

    Article  MATH  MathSciNet  Google Scholar 

  19. Fiori, V., Lintas, A., Mesrobian, S., Villa, A.: Effect of emotion and personality on deviation from purely rational decision-making. In: Guy, T., Kárný, M., Wolpert, D. (eds.) Decision Making and Imperfection. Studies in Computation Intelligence, pp. 133–164. Springer, Berlin (2013)

    Google Scholar 

  20. Fishburn, P.: Utility Theory for Decision Making. Wiley, New York (1970)

    MATH  Google Scholar 

  21. Genest, C., Zidek, J.: Combining probability distributions: a critique and annotated bibliography. Stat. Sci. 1(1), 114–148 (1986)

    Article  MathSciNet  Google Scholar 

  22. Giarlotta, A., Greco, S.: Necessary and possible preference structures. J. Math. Econ. 49, 163–172 (2013)

    Article  MATH  MathSciNet  Google Scholar 

  23. Gong, J., Zhang, Y., Yang, Z., Huang, Y., Feng, J., Zhang, W.: The framing effect in medical decision-making: a review of the literature. Psychol. Health Med. 18(6), 645–653 (2013)

    Article  Google Scholar 

  24. Grigoroudis, E., Siskos, Y.: Customer Satisfaction Evaluation: Methods for Measuring and Implementing Service Quality. International Series in Operations Research and Management. Springer, New York (2010)

    Book  Google Scholar 

  25. Guan, P., Raginsky, M., Willett, R.: Online Markov decision processes with Kullback-Leibler control cost. IEEE Trans. Autom. Control 59, 1423–1438 (2014)

    Article  MathSciNet  Google Scholar 

  26. Guy, T.V., Böhm, J., Kárný, M.: Probabilistic mixture control with multimodal target. In: Andrýsek, J., Kárný, M., Kracík, J. (eds.) Multiple Participant Decision Making, pp. 89–98. Advanced Knowledge International, Adelaide (2004)

    Google Scholar 

  27. Simon, H.A.: Models of Bounded Rationality. MacMillan, London (1997)

    Google Scholar 

  28. Haykin, S.: Neural Networks: A Comprehensive Foundation. Macmillan, New York (1994)

    MATH  Google Scholar 

  29. Insua, D., Esteban, P.: Designing societies of robots. In: Guy, T., Kárný, M. (eds.) Proceedings of the 3rd International Workshop on Scalable Decision Making Held in Conjunction with ECML/PKDD 2013. ÚTIA AVČR, Prague (2013)

    Google Scholar 

  30. Jazwinski, A.: Stochastic Processes and Filtering Theory. Academic Press, New York (1970)

    MATH  Google Scholar 

  31. Jones, B.: Bounded rationality. Annu. Rev. Polit. Sci. 2, 297–321 (1999)

    Article  Google Scholar 

  32. Kahneman, D., Tversky, A.: The psychology of preferences. Sci. Am. 246(1), 160–173 (1982)

    Article  Google Scholar 

  33. Kárný, M.: Towards fully probabilistic control design. Automatica 32(12), 1719–1722 (1996)

    Article  MATH  MathSciNet  Google Scholar 

  34. Kárný, M.: Automated preference elicitation for decision making. In: Guy, T., Kárný, M., Wolpert, D. (eds.) Decision Making and Imperfection, vol. 474, pp. 65–99. Springer, Berlin (2013)

    Chapter  Google Scholar 

  35. Kárný, M.: On approximate fully probabilistic design of decision making strategies. In: Guy, T., Kárný, M. (eds.) Proceedings of the 3rd International Workshop on Scalable Decision Making, ECML/PKDD 2013. UTIA AV ČR, Prague (2013). ISBN 978-80-903834-8-7

    Google Scholar 

  36. Kárný, M.: Approximate Bayesian recursive estimation. Inf. Sci, 289, 100–111 (2014)

    Google Scholar 

  37. Kárný, M., Andrýsek, J.: Use of Kullback-Leibler divergence for forgetting. Int. J. Adapt. Control Signal Process. 23(1), 1–15 (2009)

    Article  Google Scholar 

  38. Kárný, M., Böhm, J., Guy, T.V., Jirsa, L., Nagy, I., Nedoma, P., Tesař, L.: Optimized Bayesian Dynamic Advising: Theory and Algorithms. Springer, London (2006)

    Google Scholar 

  39. Kárný, M., Guy, T.: Preference elicitation in fully probabilistic design of  decision strategies. In: Proceedings of the 49th IEEE Conference on Decision and Control (2010)

    Google Scholar 

  40. Kárný, M., Guy, T.: Decision making with imperfect decision makers. In: Guy, T., Kárný, M., Wolpert, D. (eds.) On Support of Imperfect Bayesian Participants. Intelligent Systems Reference Library. Springer, Berlin (2012)

    Google Scholar 

  41. Kárný, M., Guy, T.V.: Fully probabilistic control design. Syst. Control Lett. 55(4), 259–265 (2006)

    Article  MATH  Google Scholar 

  42. Kárný, M., Guy, T.V., Bodini, A., Ruggeri, F.: Cooperation via sharing of probabilistic information. Int. J. Comput. Intell. Stud. 1, 139–162 (2009)

    Article  Google Scholar 

  43. Kárný, M., Halousková, A., Böhm, J., Kulhavý, R., Nedoma, P.: Design of linear quadratic adaptive control: theory and algorithms for practice. Kybernetika 21 (Supp. 3–6) (1985)

    Google Scholar 

  44. Kárný, M., Jeníček, T., Ottenheimer, W.: Contribution to prior tuning of LQG selftuners. Kybernetika 26(2), 107–121 (1990)

    MATH  MathSciNet  Google Scholar 

  45. Kárný, M., Kroupa, T.: Axiomatisation of fully probabilistic design. Inf. Sci. 186(1), 105–113 (2012)

    Article  MATH  Google Scholar 

  46. Kárný, M., Nedoma, P.: The 2nd European IEEE Workshop on Computer Intensive Methods in Control and Signal Processing. In: Berec, L., et al. (eds.) On Completion of Probabilistic Models, pp. 59–64. ÚTIA AVČR, Prague (1996)

    Google Scholar 

  47. Kerridge, D.: Inaccuracy and inference. J. R. Stat. Soc. B 23, 284–294 (1961)

    MathSciNet  Google Scholar 

  48. Knejflová, Z., Avanesyan, G., Guy, T.V., Kárný, M.: What lies beneath players’ non-rationality in ultimatum game? In: Guy, T., Kárný, M. (eds.) Proceedings of the 3rd International Workshop on Scalable Decision Making, ECML/PKDD 2013. UTIA AV ČR, Prague (2013)

    Google Scholar 

  49. Knoll, M.A.: The role of behavioral economics and behavioral decision making in Americans’ retirement savings decisions. Soc. Secur. Bull. 70(4), 1–23 (2010)

    Google Scholar 

  50. Koopman, R.: On distributions admitting a sufficient statistic. Trans. Am. Math. Soc. 39, 399 (1936)

    Article  MathSciNet  Google Scholar 

  51. Kulhavý, R.: A Bayes-closed approximation of recursive nonlinear estimation. Int. J. Adapt. Control Signal Process. 4, 271–285 (1990)

    Article  MATH  Google Scholar 

  52. Kulhavý, R., Kraus, F.J.: On duality of regularized exponential and linear forgetting. Automatica 32, 1403–1415 (1996)

    Article  MATH  Google Scholar 

  53. Kulhavý, R., Zarrop, M.B.: On a general concept of forgetting. Int. J. Control 58(4), 905–924 (1993)

    Article  MATH  Google Scholar 

  54. Kullback, S., Leibler, R.: On information and sufficiency. Ann. Math. Stat. 22, 79–87 (1951)

    Article  MATH  MathSciNet  Google Scholar 

  55. Kushner, H.: Stochastic approximation: a survey. Wiley Interdiscip. Rev. Comput. Stat. 2(1), 87–96 (2010). http://dx.doi.org/10.1002/wics.57

  56. Landa, J., Wang, X.: Bounded rationality of economic man: decision making under ecological, social, and institutional constraints. J. Bioecon. 3, 217–235 (2001)

    Article  Google Scholar 

  57. Lindley, D.: The future of statistics—a Bayesian 21st century. Suppl. Adv. Appl. Probab. 7, 106–115 (1975)

    Article  MATH  Google Scholar 

  58. Marczewski, E.: Sur l’extension de l’ordre partiel. Fundamental Mathematicae 16, 386–389 (1930). In French

    Google Scholar 

  59. McCormick, T., Raftery, A.E., Madigan, D., Burd, R.: Dynamic logistic regression and dynamic model averaging for binary classification. Technical Report Columbia University (2010)

    Google Scholar 

  60. Meditch, J.: Stochastic Optimal Linear Estimation and Control. McGraw Hill, New York (1969)

    MATH  Google Scholar 

  61. Novák, M., Böhm, J.: Adaptive LQG controller tuning. In: Hamza, M.H. (ed.) Proceedings of the 22nd IASTED International Conference Modelling, Identification and Control. Acta Press, Calgary (2003)

    Google Scholar 

  62. Novikov, A.: Optimal sequential procedures with Bayes decision rules. Kybernetika 46(4), 754–770 (2010)

    MATH  MathSciNet  Google Scholar 

  63. Peterka, V.: Bayesian system identification. In: Eykhoff, P. (ed.) Trends and Progress in System Identification, pp. 239–304. Pergamon Press, Oxford (1981)

    Chapter  Google Scholar 

  64. Pothos, E., Busemeyer, J.: A quantum probability explanation for violations of ‘rational’ decision theory. In: Proceedings of The Royal Society B, pp. 2171–2178 (2009)

    Google Scholar 

  65. Rao, M.: Measure Theory and Integration. Wiley, New York (1987)

    MATH  Google Scholar 

  66. Regenwetter, M., Dana, J., Davis-Stober, C.: Transitivity of preferences. Psychol. Rev. 118(1), 42–56 (2011)

    Article  Google Scholar 

  67. Roberts, S.: Scalable information aggregation from weak information sources. In: Guy, T., Kárný, M. (eds.) Proceedings of the 3rd International Workshop on Scalable Decision Making held in conjunction with ECML/PKDD 2013, ÚTIA AVČR, Prague (2013)

    Google Scholar 

  68. Savage, L.: Foundations of Statistics. Wiley, New York (1954)

    MATH  Google Scholar 

  69. Sečkárová, V.: On supra-Bayesian weighted combination of available data determined by Kerridge inaccuracy and entropy. Pliska Stud. Math. Bulg. 22, 159–168 (2013)

    Google Scholar 

  70. Shore, J., Johnson, R.: Axiomatic derivation of the principle of maximum entropy and the principle of minimum cross-entropy. IEEE Trans. Inf. Theory 26(1), 26–37 (1980)

    Article  MATH  MathSciNet  Google Scholar 

  71. Si, J., Barto, A., Powell, W., Wunsch, D. (eds.): Handbook of Learning and Approximate Dynamic Programming. Wiley, Danvers (2004)

    Google Scholar 

  72. Simon, H.: A behavioral model of rational choice. Q. Econ. 69, 299–310 (1955)

    Article  Google Scholar 

  73. Simon, H.: Theories of decision-making in economics and behavioral science. Am. Econ. Rev. 69, 253–283 (1959)

    Google Scholar 

  74. Syll, L.: Dutch Books, Money Pump and Bayesianism. Economics, Theory of Science and Methodology. http://larspsyll.wordpress.com/2012/06/25/dutch-books-money-pumps-and-bayesianism. (2012)

  75. Tishby, N.: Predictive information and the brain’s internal time. In: Guy, T., Kárný, M. (eds.) Proceedings of the 3rd International Workshop on Scalable Decision Making Held in  Conjunction with ECML/PKDD 2013, ÚTIA AVČR, Prague (2013)

    Google Scholar 

  76. Tishby, N., Polani, D.: Information theory of decisions and actions. In: Cutsuridis, V., Hussain, A., Taylor, J. (eds.) Perception-Action Cycle. Springer Series in Cognitive and Neural Systems, pp. 601–636. Springer, New York (2011)

    Chapter  Google Scholar 

  77. Titterington, D., Smith, A., Makov, U.: Statistical Analysis of Finite Mixtures. Wiley, New York (1985)

    Google Scholar 

  78. Todorov, E.: Advances in Neural Information Processing. In: Schölkopf, B., et al. (eds.) Linearly-solvable Markov decision problems, pp. 1369–1376. MIT Press, New York (2006)

    Google Scholar 

  79. Tordesillas, R., Chaiken, S.: Thinking too much or too little? the effects of introspection on the decision-making process. Pers. Soc. Psychol. Bull. 25, 623–629 (1999)

    Article  Google Scholar 

  80. Tversky, A., Kahneman, D.: Advances in prospect theory: cumulative representation of uncertainty. J. Risk Uncertain. 5, 297–323 (1992)

    Article  MATH  Google Scholar 

  81. Wald, A.: Statistical Decision Functions. Wiley, London (1950)

    MATH  Google Scholar 

Download references

Acknowledgments

The reported research has been supported by GAČR 13-13502S. Dr. Anthony Quinn from Trinity College, Dublin, has provided us useful feedback.

Author information

Authors and Affiliations

  1. Institute of Information Theory and Automation, The Czech Academy of Sciences, Pod Vodárenskou věží 4, 182 08, Prague 8, Czech Republic

    Miroslav Kárný & Tatiana V. Guy

Authors
  1. Miroslav Kárný
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Tatiana V. Guy
    View author publications

    You can also search for this author in PubMed Google Scholar

Corresponding author

Correspondence to Miroslav Kárný .

Editor information

Editors and Affiliations

  1. Institute of Information Theory and Automation, The Czech Academy of Sciences, Prague, Czech Republic

    Tatiana V. Guy

  2. Institute of Information Theory and Automation, The Czech Academy of Sciences, Prague, Czech Republic

    Miroslav Kárný

  3. Santa Fe Institute, Santa Fe, New Mexico, USA

    David H. Wolpert

Rights and permissions

Reprints and permissions

Copyright information

© 2015 Springer International Publishing Switzerland

About this chapter

Cite this chapter

Kárný, M., Guy, T.V. (2015). On the Origins of Imperfection and Apparent Non-rationality. In: Guy, T., Kárný, M., Wolpert, D. (eds) Decision Making: Uncertainty, Imperfection, Deliberation and Scalability. Studies in Computational Intelligence, vol 538. Springer, Cham. https://doi.org/10.1007/978-3-319-15144-1_3

Download citation

  • DOI: https://doi.org/10.1007/978-3-319-15144-1_3

  • Published:

  • Publisher Name: Springer, Cham

  • Print ISBN: 978-3-319-15143-4

  • Online ISBN: 978-3-319-15144-1

  • eBook Packages: EngineeringEngineering (R0)

Publish with us

Policies and ethics

Search

Navigation