作者drkkimo ()
看板Statistics
标题Re: 有关minimax的文章涵意?
时间Sun Sep 10 15:15:50 2006
minimax 是不是大中取小呀 是不是和决策树有关呢?
※ 引述《[email protected] ()》之铭言：
: ※ 引述《winniewo》之铭言：
: 我换个方式问好了
: > 『Bayes estimator provide a tool for solving minimax problems.
: 贝氏估计量提供了一个解决minimax问题的工具
: > Thus Bayesian considerations are helpful
: > when choosing an optimal frequentist estimator.
: 这句话看不大懂,什麽是"optimal frequentist estimator"
: > Viewed in this light, there is a
: > synthesis of the two approaches.
: 由此观点来看,就有一个两种方法的"synthesis"
: 什麽是 synthesis of the two approches.
: > The Bayesian approach provides us with a mean of
: > constructing an estimator that has
: > optimal frequentist properties.
: 贝式方法提供了我们 "with a mean of comstrcting"
: 估计量拥有 "optimal frequentist" 性质.
: > This synthesis highlights important features
: > of both the Bayesian and frequentist
: > approaches.
: 这个"sythesis"突显了贝式学与频率学重要的特性
: > The Bayesian paradigm is
: > well suited for the construction of
: > possibly optimal estimators, but is
: > less well suited for their evaluation.
: 这句的意思不太了解
: (1)什麽是construction of possibly optimal estimator
: (2)什麽是their evalustion
: (3)为什麽贝式适合的是(1)而较不适合(2)
: > The frequentist oaradigm is complementary,
: > as it is well suited for risk evaluations,
: > but less well suited for construction.
: 这个问题也跟上面一样...@@
: > It is important to view these two
: > approaches and hence "Average risk
: > optimality" and "Minimaxity and admissibility"
: > as complementary rather than adversarial;
: > together they provide a rich set of tools
: > and techniques for the statisticaian.
: 这是个对於在看这两个方法是重要的,
: 因此"平均风险最佳性"与"大中取小性和允许性"
: 为互补的,而不是相互比较的,这两个都同时提供了
: 统计学者一个好的工具与方法.
: 这是在统计教本中讲贝氏估计量与
: 大中取小估计量的一段文章,我不是学
: 这方面的东西,but考试要考啊~我也很想看懂
--
※ 发信站: 批踢踢实业坊(ptt.cc)
◆ From: 218.172.213.185