作者: Michael Zabarankin , Stan Uryasev
DOI: 10.1007/978-1-4614-8471-4_4
关键词: Combinatorics 、 Score test 、 Multivariate random variable 、 Likelihood function 、 Score 、 Estimation theory 、 Probability distribution 、 Mathematics 、 Finite set 、 M-estimator
摘要: A classical problem in the statistical decision theory is to estimate probability distribution of a random vector X given its independent observations \(x_{1},\ldots,x_{n}\). Often it assumed that comes from some family functions parametrized by set parameters \(\theta _{1},\ldots,\theta _{m}\), so this case, reduced estimating _{m}\) and called parametric estimation. However, if no specific distributions assumed, i.e., can not be completely defined finite number parameters, nonparametric