Models of Noise and Robust Estimation
作者： Federico Girosi , MASSACHUSETTS INST OF TECH CAMBRIDGE ARTIFICIAL INTELLIGENCE LAB
DOI:
关键词:
摘要: Least squares estimators are very common in statistics, but they lead to results that sensitive outliers, and it has been proposed minimize other measures of error, ``robust'''' estimates. In this paper we show using these robust corresponds assuming data corrupted by Gaussian noise whose variance fluctuates according some given probability distribution, uniquely determines the estimator.
参考文章(4)
Serge Bernstein, Sur les fonctions absolument monotones Acta Mathematica. ,vol. 52, pp. 1- 66 ,(1929) , 10.1007/BF02592679
I. J. Schoenberg, Metric spaces and completely monotone functions Annals of Mathematics. ,vol. 39, pp. 811- 841 ,(1938) , 10.2307/1968466
Davi Geiger, Federico Girosi, Parallel and deterministic algorithms from MRFs: surface reconstruction and integration european conference on computer vision. pp. 89- 98 ,(1990) , 10.1007/BFB0014854
Tomaso Poggio, Federico Girosi, Bruno Caprile, Extensions of a Theory of Networks for Approximation and Learning: Outliers and Negative Examples neural information processing systems. ,vol. 3, pp. 750- 756 ,(1990)