Skip to main content

A two-stage method for building classifiers

Abstract

A specific method for the evaluation of unknown parameters of a specified pair of normally distributed populations called the two-stage method for building classifiers is discussed. The determined parameter estimates are used to generate Bayesian data classifiers obtained by sampling from specified populations. The statistical simulation method is used to study the effect of the second evaluation stage on the mean classification error probability.

This is a preview of subscription content, access via your institution.

References

  1. 1.

    Anderson, T.W., An Introduction to Multivariete Statistical Analysis, New York: Wiley, 1958.

  2. 2.

    Rao, S.R., Linear Statistical Inference and Its Applications, New York: Wiley, 1968.

    MATH  Google Scholar 

  3. 3.

    Dinuls, R., Lorencs, A., and Mednieks, I., Performance Comparison of Methods for Tree Species Classification in Multispectral Images, Electron. Electr. Eng., 2011, no. 5(111), pp. 119–122.

  4. 4.

    Dinuls, R., Erins, G., Lorencs, A., Mednieks, I., and Sinica-Sinavskis, J., Tree Species Identification in Mixed Baltic Forest Using LiDAR and Multispectral Data, IEEE J. Selected Topics Appl. Earth Observ. Remote Sensing, 2012, vol. 5, no. 2, pp. 594–603.

    Article  Google Scholar 

Download references

Author information

Affiliations

Authors

Corresponding author

Correspondence to A. Lorencs.

Additional information

Original Russian Text © A. Lorencs, Yu. Sinitsa-Sinyavskis, 2012, published in Avtomatika i Vychislitel’naya Tekhnika, 2012, No. 5, pp. 46–57.

About this article

Cite this article

Lorencs, A., Sinitsa-Sinyavskis, Y. A two-stage method for building classifiers. Aut. Control Comp. Sci. 46, 214–222 (2012). https://doi.org/10.3103/S0146411612050045

Download citation

Keywords

  • normally distributed population
  • Bayesian classifier
  • mean classification error probability