Publications Meetings The Profession Membership Programs Math Samplings Policy & Advocacy In the News About the AMS

   
Mobile Device Pairing
Theory of Probability and Mathematical Statistics
Theory of Probability and Mathematical Statistics
ISSN 1547-7363(online) ISSN 0094-9000(print)

 

A Bayesian classifier


Authors: B. A. Zalessky and P. V. Lukashevich
Translated by: N. Semenov
Original publication: Teoriya Imovirnostei ta Matematichna Statistika, tom 78 (2008).
Journal: Theor. Probability and Math. Statist. 78 (2009), 23-35
MSC (2000): Primary 62C10; Secondary 90Bxx
Published electronically: August 4, 2009
MathSciNet review: 2446846
Full-text PDF

Abstract | References | Similar Articles | Additional Information

Abstract: We consider a new Bayesian classifier for the classification of multidimensional observations $ X_1,\dots,X_n$ of $ \mathbb{R}^k$ if the learning sample is known. We assume that the data are generated by two disjoint bounded sets $ \Omega_0,\Omega_1\subset \mathbb{R}^k$ and each vector $ X_i$ of the sample is a result of the observation after one of the sets $ \Omega_\ell$, $ \ell=0,1$, with a random error. In other words, we assume that a priori the Bayesian probability $ \mu$ is given on the set $ \Omega=\Omega_0\cup\Omega_1$ and that every vector of observations $ X_i$ has the density

$\displaystyle g_\ell(x)=q_\ell\int_{\Omega_\ell}f(x,y) \mu(dy),\qquad \ell=0,1, $

where the function $ f(x,y)$ is a probability density for all $ y\in \Omega$ and $ q_\ell^{-1}=\mu(\Omega_\ell)$.

The maximum a posteriori probability estimators $ \widehat{\Omega}_{\ell,n}$, $ \ell=0,1$, for the sets  $ \Omega_\ell$, $ \ell=0,1$, are constructed with the help of the learning sample. Under natural assumptions imposed on $ \Omega_0$ and $ \Omega_1$, we show that the estimators converge to some sets (possibly different from $ \Omega_0$ and $ \Omega_1$). If the mean frequencies $ \pi_\ell$ of observations of the classes $ \Omega_\ell$ are equal to $ \mu(\Omega_\ell)$, $ \ell=0,1$, then the estimators are consistent in the sense that $ \widehat{\Omega}_{\ell,n} \stackrel{n\rightarrow\infty}{\longrightarrow}\Omega_{\ell}$, $ \ell=0,1$. We also discuss some results of numerical experiments showing the applicability of our classifier for solving the problems of the statistical classification.


References [Enhancements On Off] (What's this?)

  • 1. S. A. Aivazyan, B. M. Buchshtaber, I. S. Enyukov, and L. D. Meshalkin, Applied Statistics: Classification and Reducing of Dimension, Finansy i Statistika, Moscow, 1989. (Russian)
  • 2. A. A. Borovkov, Matematicheskaya statistika, “Nauka”, Moscow, 1984 (Russian). Otsenka parametrov. Proverka gipotez. [Estimation of parameters. Testing of hypotheses]. MR 782295 (86i:62001)
  • 3. Leo Breiman, Jerome H. Friedman, Richard A. Olshen, and Charles J. Stone, Classification and regression trees, Wadsworth Statistics/Probability Series, Wadsworth Advanced Books and Software, Belmont, CA, 1984. MR 726392 (86b:62101)
  • 4. L. Breiman, Random Forests, Technical report, Department of Statistics, University of California, Berkeley, CA, 1999.
  • 5. Vladimir Vapnik, Estimation of dependences based on empirical data, Springer Series in Statistics, Springer-Verlag, New York-Berlin, 1982. Translated from the Russian by Samuel Kotz. MR 672244 (84a:62043)
  • 6. Vladimir N. Vapnik, Statistical learning theory, Adaptive and Learning Systems for Signal Processing, Communications, and Control, John Wiley & Sons, Inc., New York, 1998. A Wiley-Interscience Publication. MR 1641250 (99h:62052)
  • 7. S. Haykin, Neural Networks: A Comprehensive Foundation, Wiley, New York, 2005.
  • 8. E. E. Zhuk and Yu. S. Kharin, Stability in the Cluster Analysis of Multivariate Data, Belgosuniversitet, Minsk, 1998. (Russian)
  • 9. Shelemyahu Zacks, The theory of statistical inference, John Wiley & Sons, Inc., New York-London-Sydney, 1971. Wiley Series in Probability and Mathematical Statistics. MR 0420923 (54 #8934a)
  • 10. È. Leman, Teoriya tochechnogo otsenivaniya, \cyr Teoriya Veroyatnosteĭ\ i Matematicheskaya Statistika [Probability Theory and Mathematical Statistics], vol. 43, “Nauka”, Moscow, 1991 (Russian). Translated from the English by Yu.\ V. Prokhorov. MR 1143059 (93c:62003b)
  • 11. G. Matheron, Random sets and integral geometry, John Wiley\thinspace&\thinspace Sons, New York-London-Sydney, 1975. With a foreword by Geoffrey S. Watson; Wiley Series in Probability and Mathematical Statistics. MR 0385969 (52 #6828)
  • 12. V. V. Mottl′ and I. B. Muchnik, Skrytye markovskie modeli v strukturnom analize signalov, Fiziko-Matematicheskaya Literatura, Moscow, 1999 (Russian, with Russian summary). MR 1778152 (2001m:94014)
  • 13. J. Pfanzagl, On the measurability and consistency of minimum contrast estimates, Metrika 14 (1969), 249-273.
  • 14. D. Forsyth and J. Ponce, Computer Vision. A Modern Approach, Prentice Hall, New York, 2002.
  • 15. Keinosuke Fukunaga, Introduction to statistical pattern recognition, 2nd ed., Computer Science and Scientific Computing, Academic Press, Inc., Boston, MA, 1990. MR 1075415 (91i:68131)
  • 16. M. I. Schlesinger and V. Hlavac, Ten Lectures on Statistical and Structural Pattern Recognition, Springer-Verlag, Berlin, 2002.

Similar Articles

Retrieve articles in Theory of Probability and Mathematical Statistics with MSC (2000): 62C10, 90Bxx

Retrieve articles in all journals with MSC (2000): 62C10, 90Bxx


Additional Information

B. A. Zalessky
Affiliation: United Institute of Informatics Problems, National Academy of Sciences, Surganova Street 6, Minsk, 220012, Belarus’
Email: zalesky@newman.bas-net.by

P. V. Lukashevich
Affiliation: United Institute of Informatics Problems, National Academy of Sciences, Surganova Street 6, Minsk, 220012, Belarus’

DOI: http://dx.doi.org/10.1090/S0094-9000-09-00759-5
PII: S 0094-9000(09)00759-5
Received by editor(s): October 23, 2006
Published electronically: August 4, 2009
Additional Notes: The first author was supported by the INTAS grant 04-77-7036
Article copyright: © Copyright 2009 American Mathematical Society