由买买提看人间百态

boards

本页内容为未名空间相应帖子的节选和存档,一周内的贴子最多显示50字,超过一周显示500字 访问原贴
Statistics版 - recommended reading
相关主题
如何同时测试2000组数据是否正太分布Best test of normality?
normality checkRe: 统计经典教科书 (we have so far)
Paired T-test Assumption not Satisfied[合集] 如何检验normality??
问2个统计检验的问题how to determine data fit some distribution? thanks
请推荐一个data mining 的书。发包子请人review我的syllabus - stochastics
[转载] 请问大家考精算都用什么样的STUDY MANUALwhich nonparametric test shoud I do
Bidirectional Von Mises DistributionKolmogorov-Smirnov test for discrete distributions
如何test两个Poisson分布的mean difference? 急有关于cross validation的东西一般出现在哪方面的统计书中?
相关话题的讨论汇总
话题: robert话题: burdzy话题: he话题: his话题: bayesian
进入Statistics版参与讨论
1 (共1页)
s*******d
发帖数: 126
1
in light of what is happening on this board, per discussions/criticisms on
Chen and his idea(s), I am reminded of this review paper that was published
in "Bayesian Analysis" last year,
"The Search for Certainty: A critical assessment" by Christian Robert
http://ba.stat.cmu.edu/journal/2010/vol05/issue02/robert.pdf
Here is some background information, which, to some extend, mirros what's
happening on this board.
Krzysztof Burdzy is a professor from U. of Washington and he wrote a book
called "The Search for Certainty", which, among other things, critizied two
very well established views on the foundation of probabiilty; de Finetti's
subjective probability (Bayesian) and von Mises's frequency-based approach.
(You don't need to understand any of this, or even be interested in them, to
read this review article by Robert). And, check this out, Burdzy not only
came out and said that de finetti and von Mises were inadequate, he had the
audacity to come up with his OWN version of the probability axioms to
replace Kolmogorov's axioms! Can you imagine this, in this day and age?! Isn
't that just crazy?
Apparently this really pissed Robert off (Robert is a fairly well known
Bayesian statistician from France). So he wrote this review paper, taking
apart Burdzy's book (Burdzy is not a statistician by training, and he is
more of a math/philosophy type of guy, so the language, or the (
unconventional) usage of it, didn't please Robert much neither).
I am not remotely close to suggesting that Chen is Burdzy, and I am not here
to defend him. But I have to admit that this Robert's paper is a facinate
read because you can tell that he is completely pissed off by this Burdzy
dude but he has to be critical in a rigrous way. His paper shows a good
example on how these academic conversations on a high-level should be
conducted. And this is something that everyone on this board can learn
something from.
And what's even more facinating is one of the comment papers by Larry
Wasserman, a professor from Carnegie Mellon University http://ba.stat.cmu.edu/journal/2010/vol05/issue02/wasserman.pdf. He was able to see through Burdzy's arguments and came up with his own understanding and delivered his opinions in a very succinct fashion that points to the heart of the problem.
So if you have a few minutes of your time, I highly recommand reading Robert
's paper and Wasserman's comment. These are quite short and non-technical
papers that are only a few pages long. Plus, the way they argue is what you
want to read, not necessarily what is being discussed.
Enjoy!
p.s. link to the whole issue is http://ba.stat.cmu.edu/vol05is02.php
D******n
发帖数: 2836
2
downloaded.

published
two

【在 s*******d 的大作中提到】
: in light of what is happening on this board, per discussions/criticisms on
: Chen and his idea(s), I am reminded of this review paper that was published
: in "Bayesian Analysis" last year,
: "The Search for Certainty: A critical assessment" by Christian Robert
: http://ba.stat.cmu.edu/journal/2010/vol05/issue02/robert.pdf
: Here is some background information, which, to some extend, mirros what's
: happening on this board.
: Krzysztof Burdzy is a professor from U. of Washington and he wrote a book
: called "The Search for Certainty", which, among other things, critizied two
: very well established views on the foundation of probabiilty; de Finetti's

g********r
发帖数: 8017
3
reading。I highly suspect 大师 will read it too, and twist the terms to use
for his technically false arguments shrouded by ambiguity.
T*******I
发帖数: 5138
4
I almost miss this. Thanks very much. I will take a look.

published
two

【在 s*******d 的大作中提到】
: in light of what is happening on this board, per discussions/criticisms on
: Chen and his idea(s), I am reminded of this review paper that was published
: in "Bayesian Analysis" last year,
: "The Search for Certainty: A critical assessment" by Christian Robert
: http://ba.stat.cmu.edu/journal/2010/vol05/issue02/robert.pdf
: Here is some background information, which, to some extend, mirros what's
: happening on this board.
: Krzysztof Burdzy is a professor from U. of Washington and he wrote a book
: called "The Search for Certainty", which, among other things, critizied two
: very well established views on the foundation of probabiilty; de Finetti's

1 (共1页)
进入Statistics版参与讨论
相关主题
有关于cross validation的东西一般出现在哪方面的统计书中?请推荐一个data mining 的书。
qqplot和非参检验[转载] 请问大家考精算都用什么样的STUDY MANUAL
怎么区分正态和T分布?Bidirectional Von Mises Distribution
问个goodness of fit的问题如何test两个Poisson分布的mean difference? 急
如何同时测试2000组数据是否正太分布Best test of normality?
normality checkRe: 统计经典教科书 (we have so far)
Paired T-test Assumption not Satisfied[合集] 如何检验normality??
问2个统计检验的问题how to determine data fit some distribution? thanks
相关话题的讨论汇总
话题: robert话题: burdzy话题: he话题: his话题: bayesian