Abstract
In the last decade, new methods of forecasting were developed different from traditional statistical methods. In particular, it is possible to “efficiently” predict any sequence of outcomes without using any hypothesis on the nature of a source generating it. In the present paper, a modified version of the universal forecasting algorithm is considered. The main part of the paper is devoted to algorithmic analysis of universal forecasting methods and to exploring limits of their performance.
Similar content being viewed by others
References
V’yugin V.V. On Sequences with Non-learnable Subsequences, Computer Science-Theory and Applications (Proc. 3rd Int. Computer Science Sympos. in Russia [CSR’2008], Moscow, Russia, 2008), Hirsh, E., Razborov, A., Semenov, A., and Slissenko, A., Eds., Lect. Notes Comp. Sci., vol. 5010, Berlin: Springer, 2008, pp. 302–313.
Dawid, A.P., TheWell-Calibrated Bayesian, J. Amer. Statist. Assoc., 1982, vol. 77, no. 379, pp. 605–613.
Foster, D.P. and Vohra, R.V., Asymptotic Calibration, Biometrika, 1998, vol. 85, no. 2, pp. 379–390.
Sandroni, A., The Reproducible Properties of Correct Forecasts, Int. J. Game Theory, 2003, vol. 32, no. 1, pp. 151–159.
Dekel, E. and Feinberg, Y., Non-Bayesian Testing of a Stochastic Prediction, Rev. Econom. Stud., 2006, vol. 73, no. 4, pp. 893–906.
Fortnow, L. and Vohra, R.V., The Complexity of Forecast Testing, Econometrica, 2009, vol. 77, no. 1, pp. 93–105.
Dawid, A.P., Calibration-Based Empirical Probability, Ann. Statist., 1985, vol. 13,no. 4, pp. 1251–1285.
Uspensky, V.A., Semenov, A.L., and Shen’, A.Kh., Can an Individual Sequence of Zeros and Ones Be Random?, Uspekhi Mat. Nauk, 1990, vol. 45, no. 1, pp. 105–162 [Russian Math. Surveys (Engl. Transl.), 1990, vol. 45, no. 1, pp. 121–189].
Li, M. and Vitányi, P., An Introduction to Kolmogorov Complexity and Its Applications, New York: Springer, 1997, 2nd ed.
Oakes, D., Self-Calibrating Priors Do Not Exist, J. Amer. Statist. Assoc., 1985, vol. 80, no. 390, pp. 339–342.
Sandroni, A., Smorodinsky, R., and Vohra, R.V., Calibration with Many Checking Rules, Math. Oper. Res., 2003, vol. 28, no. 1, pp. 141–153.
Shiryaev, A.N., Veroyatnost’, Moscow: Nauka, 1980. Translated under the title Probability, New York: Springer, 1984.
Kakade, S. and Foster, D.P., Deterministic Calibration and Nash Equilibrium, Proc. 17th Ann. Conf. on Learning Theory (COLT’2004), Banff, Canada, 2004, Shawe-Taylor, J. and Singer, Y., Eds., Lect. Notes Comp. Sci., vol. 3120, Berlin: Springer, 2004, pp. 33–48.
Lehrer, E., Any Inspection Rule Is Manipulable, Econometrica, 2001, vol. 69, no. 5, pp. 1333–1347.
Rogers, H., Theory of Recursive Functions and Effective Computability, New York: McGraw-Hill, 1967. Translated under the title Teoriya rekursivnykh funktsii i effektivnaya vychislimost’, Moscow: Mir, 1972.
Zvonkin, A.K. and Levin, L.A., Complexity of Finite Objects and the Algorithmic Concepts of Information and Randomness, Uspekhi Mat. Nauk, 1970, vol. 25, no. 6, pp. 85–127 [Russian Math. Surveys (Engl. Transl.), 1970, vol. 25, no. 6, pp. 83–124].
V’yugin, V.V., Non-Stochastic Infinite and Finite Sequences, Theoret. Comput. Sci., 1998, vol. 207, no. 2, pp. 363–382.
Vovk, V., Defensive Forecasting for Optimal Prediction with Expert Advice, arXiv e-print 0708.1503v1, 2007.
Cesa-Bianchi, N. and Lugosi, G., Prediction, Learning, and Games, Cambridge: Cambridge Univ. Press, 2006.
Author information
Authors and Affiliations
Corresponding author
Additional information
Original Russian Text © V.V. V’yugin, 2011, published in Problemy Peredachi Informatsii, 2011, Vol. 47, No. 2, pp. 90–116.
The paper is an extended version of a talk [1] given at the Computer Science Symposium in Russia (CSR-2008), Moscow, 2008.
Supported in part by the Russian Foundation for Basic Research, project nos. 09-07-00180a and 09-01-00709a.
Rights and permissions
About this article
Cite this article
V’yugin, V.V. On universal algorithms for adaptive forecasting. Probl Inf Transm 47, 166–189 (2011). https://doi.org/10.1134/S0032946011020074
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1134/S0032946011020074