DocumentCode :
2621748
Title :
Nonparametric estimation using neural networks
Author :
Lugosi, Gábor ; Zeger, Kenneth
Author_Institution :
Dept. of Math., Budapest Univ., Hungary
fYear :
1994
fDate :
27 Jun-1 Jul 1994
Firstpage :
112
Abstract :
We show that properly trained neural networks provide universally consistent nonparametric estimators. The results apply to regression estimation, conditional median estimation, curve fitting, pattern recognition and learning concepts. The estimators minimize the empirical Lp-error
Keywords :
curve fitting; learning (artificial intelligence); neural nets; pattern recognition; statistical analysis; conditional median estimation; curve fitting; learning; neural networks; nonparametric estimation; pattern recognition; regression estimation; Approximation error; Convergence; Estimation error; Feedforward neural networks; Mathematics; Neural networks; Neurons; Pattern recognition; Random variables; Risk management;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Information Theory, 1994. Proceedings., 1994 IEEE International Symposium on
Conference_Location :
Trondheim
Print_ISBN :
0-7803-2015-8
Type :
conf
DOI :
10.1109/ISIT.1994.394876
Filename :
394876
Link To Document :
بازگشت