Finiteness of the V γ dimension of the set of Huber loss function for regression is studied in a ball of radius R in an infinite dimension RKHS,and an upper bound of it is estimated.

In order to estimate the background motion using only the background pixels and reduce the influence of moving objects to a great extend, Huber function is considered to discriminate between the.

Mis-acceptance and mis-rejection in product examination and analysis of quality loss function;

Research on stochastic facility layout problem based on loss function;

Multi-output support vector regression with piecewise loss function;

The sample size in random sampling can then be determined through forming the lost function.

Bayesian estimation of geometric distribution parameterunder entropy loss function;

In this paper,the formula of Expect Bayes estimation of the reliability under entropy loss function for geometric distribution have been given,when the prior distribution of the reliability is power distribution and βdistribution.

This paper considers comparison of MINQUE and simple estimator of Σ in the mult-ivariate normal linear model Y-N(XB, Σ V) under the risk of entropy loss function and symmetry loss function criterion, where the design matrix X need not have full rank and the dispersion matrix V can be singular.

Under Linex loss function L(θ,δ)=ec(δ-θ)-c(δ-θ)-1,c>0,It is proved that the unique Bayes estimator δB(x),the multilayer Bayes estimator δ^B(X) and the general form of the admissible estimator are δB(X)=-1cln E(e-cθ|X)=n+αcln(1+cλ+T),δ^B(X)=-1cln(∫c0∫10Kλα(λ+c+T)n+αdαdλ∫c0∫10Kλα(λ+T)n+αdαdλ) and Sln(1+cd+T) respectively for the scale parameter reciprocal of the Rayleigh distribution.

First, we obtain the empirical Bayesian estimator of the scale parameter for the double exponential distribution based on LINEX loss function and the convergence rate of the estimate.