论文信息 - On Bayes Risk Consistent Pattern Recognition Procedures in a Quasi-Stationary Environment

On Bayes Risk Consistent Pattern Recognition Procedures in a Quasi-Stationary Environment

Van Ryzin and Greblicki showed that pattern recognition procedures derived from orthogonal series estimates of a probability density function are Bayes risk consistent. In this note it is proved that these procedures do not lose-under some additional conditions-their asymptotic properties even if the random environment is nonstationary.

Leszek Rutkowski

[1] Wlodzimierz Greblicki,et al. Asymptotically optimal pattern recognition procedures with density estimates (Corresp.) , 1978, IEEE Trans. Inf. Theory.

[2] G. Alexits,et al. Convergence Problems of Orthogonal Series. , 1961 .

[3] T. Wagner,et al. ASYMPTOTICALLY OPTIMAL DISCRIMINANT FUNCTIONS FOR PATTERN RECOGNITION. , 1968 .

[4] T. Wagner,et al. Asymptotically optimal discriminant functions for pattern classification , 1969, IEEE Trans. Inf. Theory.

[5] S. Schwartz. Estimation of Probability Density by an Orthogonal Series , 1967 .

[6] Azriel Rosenfeld,et al. Neighbor gray levels as features in pixel classification , 1980, Pattern Recognit..

[7] L. Devroye,et al. On the L1 convergence of kernel estimators of regression functions with applications in discrimination , 1980 .

[8] Terry J. Wagner,et al. Convergence of the nearest neighbor rule , 1970, IEEE Trans. Inf. Theory.