>>973
追加
Alexander Prussさん
(抜粋)
Assume CH. Let < be a well-order of [0,1]. Suppose X and Y are i.i.d. uniformly distributed on [0,1].
By a conglomerability assumption, we could then conclude that P(X <= Y)=0, which would be absurd as the same reasoning would also show that P(Y <= X)=0.
The argument fallaciously assumes conglomerability.
We are neither justified in concluding that P(X <= Y)=0, nor that {X <= Y} is measurable (though for each fixed y, {X <= y} is measurable).
And indeed it's not measurable: for were it measurable, we could use Fubini to conclude that it has null probability.
Note that one can repeat the argument without CH but instead using an extension of Lebesgue measure that assigns null probability to every subset of cardinality < Nc, so clearly there is no refutation of CH here.
(CH： Continuum Hypothesis )
(引用終り)

この議論は、私スレ主が、確率論の専門家さんと呼ぶ人の議論とほぼ同じですね
時枝先生は、記事中で、ビタリ類似をもって、非可測としているが、Alexander Prussさんと、確率論の専門家さんとは、
別の視点から、”not measurable”あるいは”null probability”だという
なお、上記冒頭で、”i.i.d. ”が登場していることを注意しておきます（＾＾；

スレ20 http://rio2016.2ch.net/test/read.cgi/math/1466279209/528-529
528 名前：１３２人目の素数さん[] 投稿日：2016/07/03(日) 23:03:57.29 ID:f9oaWn8A [8/13]
おれが問題視してるのはの可測性
正確にかくために確率空間(Ω,F,P)を設定しよう
Y,Zはそれぞれ(Ω,F)から(R,B(R))の可測関数である．
もしhが(R,B(R))から(N,2^N)への可測関数ならば
h(Y),h(Z)はそれぞれ可測関数となって{ω|h(Y(ω))>h(Z(ω)}∈FとなりP({ω|h(Y(ω))>h(Z(ω)})=1/2となるけど
hが(R,B(R))から(N,2^N)への可測関数とは正直思えない

529 名前：１３２人目の素数さん[] 投稿日：2016/07/03(日) 23:04:46.18 ID:f9oaWn8A [9/13]
>>528
自己レス
(R,B(R))ではなくすべて(R^N,B(R^N))だな
(引用終り)