Skip to main content
SpringerLink
Log in
Book cover

Introduction to Artificial Intelligence pp 113–160Cite as

Reasoning with Uncertainty

  • Chapter

  • 12k Accesses

Part of the book series: Undergraduate Topics in Computer Science ((UTICS))

Abstract

Reasoning under uncertainty with limited resources and incomplete knowledge plays a big role in everyday situations and also in many technical applications of AI. Probabilistic reasoning is the modern AI method for solving these problems. After a brief introduction to probability theory we present the powerful method of maximum entropy and Bayesian networks which are used in many applications. The medical diagnosis expert system LEXMED, developed by the author, is used to demonstrate the power of these formalisms.

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

Notes

  1. 1.

    The computed probabilities can only be used for continued propositions if the measured sample (100 vehicles) is representative. Otherwise only propositions about the observed 100 vehicles can be made.

  2. 2.

    A set of probabilistic equations is called consistent if there is at least one solution, that is, one distribution which satisfies all equations.

  3. 3.

    The reader may calculate this result by maximization of the entropy under the normalization condition (Exercise 7.5 on page 158).

  4. 4.

    QP([penguin=yes]-|> [flies=yes]) is an alternative form of the PIT syntax for QP([flies=yes] | [penguin=yes]).

  5. 5.

    The project was financed by the German state of Baden-Württemberg, the health insurance company AOK Baden-Württemberg, the Ravensburg-Weingarten University of Applied Sciences, and the 14 Nothelfer Hospital in Weingarten.

  6. 6.

    These negative diagnoses are denoted “non-specific abdominal pain” (NSAP).

  7. 7.

    The task of generating a function from a set of data is known as machine learning. We will cover this thoroughly in Chap. 8.

  8. 8.

    A version with limited functionality is accessible without a password.

  9. 9.

    Instead of individual numerical values, intervals can also be used here (for example [0.06,0.12]).

  10. 10.

    For a systematic introduction to machine learning we refer the reader to Chap. 8.

  11. 11.

    The difference between this and a Bayesian network is, for example, that the rules are equipped with probability intervals and that only after applying the principle of maximum entropy is a unique probability model produced.

  12. 12.

    Ambulant observation means that the patient is released to stay at home.

  13. 13.

    In the naive Bayes method, the independence of all attributes is assumed, and this method has been successfully applied to text classification (see Sect. 8.6).

  14. 14.

    The binary variables J and M stand for the two events “John calls”, and “Mary calls”, respectively, Al for “alarm siren sounds”, Bur for “burglary” and Ear for “earthquake”.

  15. 15.

    For the case of a node without ancestors the product in this sum is empty. For this we substitute the value 1 because the CPT for nodes without ancestors contains, with its a priori probability, exactly one value.

  16. 16.

    If for example three nodes X 1, X 2, X 3 form a cycle, then there are the edges (X 1,X 2), (X 2,X 3) and (X 3,X 1) where X 1 has X 3 as a successor.

  17. 17.

    This is also not always quite so simple.

  18. 18.

    In Sect. 8.6 and in Exercise 8.16 on page 219 we will show that the scores are equivalent to the special case naive Bayes, that is, to the assumption that all symptoms are conditionally independent given the diagnosis.

References

  1. E. W. Adams. The Logic of Conditionals. Synthese Library, volume 86. Reidel, Dordrecht, 1975.

    MATH  Google Scholar 

  2. P. Cheeseman. A method for computing generalised bayesian probability values for expert systems. In Proc. of the 8th Intl. Joint Conf. on Artificial Intelligence (IJCAI-83), 1983.

    Google Scholar 

  3. P. Cheeseman. In defense of probability. In Proc. of the 9th Intl. Joint Conf. on Artificial Intelligence (IJCAI-85), 1985.

    Google Scholar 

  4. F. G. Cozman. Javabayes, bayesian networks in java, 1998. www.cs.cmu.edu/~javabayes.

  5. F. T. de Dombal. Diagnosis of Acute Abdominal Pain. Churchill Livingstone, London, 1991.

    Google Scholar 

  6. F. T. de Dombal, D. J. Leaper, J. R. Staniland, A. P. McCann, and J. C. Horrocks. Computer aided diagnosis of acute abdominal pain. Br. Med. J., 2:9–13, 1972.

    Article  Google Scholar 

  7. R. O. Duda, P. E. Hart, and D. G. Stork. Pattern Classification. Wiley, New York, 2001.

    MATH  Google Scholar 

  8. C. Elkan. The paradoxical success of fuzzy logic. In Proceedings of the Eleventh National Conference on Artificial Intelligence (AAAI-93), pages 698–703. MIT Press, Cambridge, 1993.

    Google Scholar 

  9. W. Ertel. Artificial Intelligence, 2011. www.hs-weingarten.de/~ertel/aibook. Homepage to this book with materials, demo programs, links, literature, errata, etc.

  10. W. Ertel and M. Schramm. Combining data and knowledge by MaxEnt-optimization of probability distributions. In PKDD’99 (3rd European Conference on Principles and Practice of Knowledge Discovery in Databases). LNCS, volume 1704, pages 323–328. Springer, Prague, 1999.

    Chapter  Google Scholar 

  11. D. Freedman, R. Pisani, and R. Purves. Statistics. Norton, New York, 4th edition, 2007.

    Google Scholar 

  12. B. Hontschik. Theorie und Praxis der Appendektomie. Mabuse, Frankfurt am Main, 1994.

    Google Scholar 

  13. E. T. Jaynes. Information theory and statistical mechanics. Phys. Rev., 1957.

    Google Scholar 

  14. E. T. Jaynes. Probability Theory: The Logic of Science. Cambridge University Press, Cambridge, 2003.

    Book  MATH  Google Scholar 

  15. F. V. Jensen. Bayesian Networks and Decision Graphs. Springer, Berlin, 2001.

    MATH  Google Scholar 

  16. Th. Kane. Maximum entropy in Nilsson’s probabilistic logic. In Proc. of the 11th Intl. Joint Conf. on Artificial Intelligence (IJCAI-89), 1989.

    Google Scholar 

  17. J. N. Kapur and H. K. Kesavan. Entropy Optimization Principles with Applications. Academic Press, San Diego, 1992.

    Google Scholar 

  18. Lexmed—a learning expert system for medical diagnosis, 1999. www.lexmed.de.

  19. N. J. Nilsson. Probabilistic logic. Artif. Intell., 28(1):71–87, 1986.

    Article  MATH  MathSciNet  Google Scholar 

  20. C. Ohmann, C. Franke, Q. Yang, M. Margulies, M. Chan, van P. J. Elk, F. T. de Dombal, and H. D. Röher. Diagnosescore für akute Appendizitis. Chirurg, 66:135–141, 1995.

    Google Scholar 

  21. C. Ohmann, V. Moustakis, Q. Yang, and K. Lang. Evaluation of automatic knowledge acquisition techniques in the diagnosis of acute abdominal pain. Artif. Intell. Med., 8:23–36, 1996.

    Article  Google Scholar 

  22. C. Ohmann, C. Platen, and G. Belenky. Computerunterstütze Diagnose bei akuten Bauchschmerzen. Chirurg, 63:113–123, 1994.

    Google Scholar 

  23. J. Pearl. Probabilistic Reasoning in Intelligent Systems. Networks of Plausible Inference. Morgan Kaufmann, San Mateo, 1988.

    Google Scholar 

  24. W. Rödder and C.-H. Meyer. Coherent knowledge processing at maximum entropy by SPIRIT. In KI-96 (German National Conference on AI), Dresden, 1996.

    Google Scholar 

  25. S. Russell and P. Norvig. Artificial Intelligence: A Modern Approach. Prentice Hall, New York, 3rd edition, 2010. 1st edition: 1995. http://aima.cs.berkeley.edu.

    Google Scholar 

  26. S. M. Ross. Introduction to Probability and Statistics for Engineers and Scientists. Academic Press, San Diego, 2009.

    MATH  Google Scholar 

  27. M. Schramm. Indifferenz, Unabhängigkeit und maximale Entropie: Eine wahrscheinlichkeitstheoretische Semantik für Nicht-Monotones Schließen. Dissertationen zur Informatik, Band 4. CS, Munich, 1996.

    MATH  Google Scholar 

  28. M. Schramm and W. Ertel. Reasoning with probabilities and maximum entropy: the system PIT and its application in LEXMED. In K. Inderfurth et al., editor, Operations Research Proceedings (SOR’99), pages 274–280. Springer, Berlin, 2000.

    Google Scholar 

  29. E. H. Shortliffe. Computer-Based Medical Consultations, MYCIN. North-Holland, New York, 1976.

    Google Scholar 

  30. J. Stewart. Multivariable Calculus. Brooks Cole, Florence, 2007.

    Google Scholar 

  31. C. E. Shannon and W. Weaver. Mathematische Grundlagen der Informationstheorie. Oldenbourg, Munich, 1976.

    MATH  Google Scholar 

  32. J. Whittaker. Graphical Models in Applied Multivariate Statistics. Wiley, New York, 1996.

    Google Scholar 

  33. A. Zielke, H. Sitter, T. A. Rampp, E. Schäfer, C. Hasse, W. Lorenz, and M. Rothmund. Überprüfung eines diagnostischen Scoresystems (Ohmann-Score) für die akute Appendizitis. Chirurg, 70:777–783, 1999.

    Article  Google Scholar 

Download references

Author information

Authors and Affiliations

  1. FB Elektrotechnik und Informatik, Hochschule Ravensburg-Weingarten, University of Applied Sciences, Weingarten, Germany

    Prof. Dr. Wolfgang Ertel

Authors
  1. Prof. Dr. Wolfgang Ertel
    View author publications

    You can also search for this author in PubMed Google Scholar

Corresponding author

Correspondence to Wolfgang Ertel .

Rights and permissions

Reprints and permissions

Copyright information

© 2011 Springer-Verlag London Limited

About this chapter

Cite this chapter

Ertel, W. (2011). Reasoning with Uncertainty. In: Introduction to Artificial Intelligence. Undergraduate Topics in Computer Science. Springer, London. https://doi.org/10.1007/978-0-85729-299-5_7

Download citation

  • DOI: https://doi.org/10.1007/978-0-85729-299-5_7

  • Publisher Name: Springer, London

  • Print ISBN: 978-0-85729-298-8

  • Online ISBN: 978-0-85729-299-5

  • eBook Packages: Computer ScienceComputer Science (R0)

Publish with us

Policies and ethics

Search

Navigation