勘误表:具有常见噪声的平均场游戏

IF 2.1 1区 数学 Q1 STATISTICS & PROBABILITY
R. Carmona, F. Delarue, D. Lacker
{"title":"勘误表:具有常见噪声的平均场游戏","authors":"R. Carmona, F. Delarue, D. Lacker","doi":"10.1214/20-aop1432","DOIUrl":null,"url":null,"abstract":"This note corrects Lemma 3.7 in our paper [1]. The main results of the paper remain correct as stated. This note corrects an error in [1, Lemma 3.7]. The lemma is not correct as stated, and the first conclusion must instead be stated as a hypothesis. This erratum corrects the statement of the lemma and then shows that the additional hypothesis is satisfied in each of the three applications of the lemma later in the paper. The main results of the paper remain unchanged. The error in [1, Lemma 3.7] is at the end of the “first step” of the proof. Specifically, the last equation before the “second step” (lines 5-6 of page 3769) is not accurate, because the preceding equation was proven only for all F t -measurable functions φ1(μ), not for all F μ T -measurable functions. We rewrite the lemma as follows, stating equivalence between its two claims as well as a third and often more convenient form: Lemma 3.7*. Let P ∈ Pp(Ω) such that (B,W ) is a Wiener process with respect to the filtration (F t )t∈[0,T ] under P , and define ρ := P ◦ (ξ,B,W, μ)−1. Suppose that (1) and (3) of Definition 3.4 are satisfied, that P (X0 = ξ) = 1, and that the state equation (3.3) holds under P . The following are equivalent: (A) For P ◦ μ−1-almost every ν ∈ Pp(X ), it holds that (Wt)t∈[0,T ] is an (F W,Λ,X t )t∈[0,T ]-Wiener process under ν. (B) Under P , F T ∨ F ξ,W,Λ t is independent of σ{Ws −Wt : s ∈ [t, T ]} for every t ∈ [0, T ). (C) P is an MFG pre-solution Proof. (A⇒ C): Let Q = P ◦ (ξ,B,W, μ,Λ)−1. Assuming (A) holds, the second and third steps of the original proof [1, Lemma 3.7] are correct and show that Q ∈ A(ρ). As all of the other defining properties of an MFG pre-solution hold by assumption, we deduce (C). (C ⇒ B): Note that (C) entails that FΛ t is conditionally independent of F ξ,B,W,μ T given F ξ,B,W,μ t under P , for every t ∈ [0, T ). Fix t ∈ [0, T ), and fix bounded functions φt, ψT , ψt, and ht+ such that φt(Λ) is FΛ t -measurable, ψT (B,μ) is F B,μ T -measurable, ψt(ξ,W ) is F ξ,W t -measurable, and ht (W ) is σ{Ws −Wt : s ∈ [t, T ]}-measurable. The conditional independence yields E [ φt(Λ)| F T ] = E [ φt(Λ)| F t ] , a.s. The independence of ξ, (B,μ), and W easily implies that F T ∨F ξ,W t is independent of σ{Ws− Wt : s ∈ [t, T ]}, and we deduce E [φt(Λ)ψT (B,μ)ψt(ξ,W )ht+(W )] = E [ E [ φt(Λ)| F t ] ψT (B,μ)ψt(ξ,W )ht+(W ) ] = E [ E [ φt(Λ)| F t ] ψT (B,μ)ψt(ξ,W ) ] E [ht+(W )] = E [φt(Λ)ψT (B,μ)ψt(ξ,W )]E [ht+(W )] .","PeriodicalId":50763,"journal":{"name":"Annals of Probability","volume":null,"pages":null},"PeriodicalIF":2.1000,"publicationDate":"2020-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"Errata: Mean field games with common noise\",\"authors\":\"R. Carmona, F. Delarue, D. Lacker\",\"doi\":\"10.1214/20-aop1432\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"This note corrects Lemma 3.7 in our paper [1]. The main results of the paper remain correct as stated. This note corrects an error in [1, Lemma 3.7]. The lemma is not correct as stated, and the first conclusion must instead be stated as a hypothesis. This erratum corrects the statement of the lemma and then shows that the additional hypothesis is satisfied in each of the three applications of the lemma later in the paper. The main results of the paper remain unchanged. The error in [1, Lemma 3.7] is at the end of the “first step” of the proof. Specifically, the last equation before the “second step” (lines 5-6 of page 3769) is not accurate, because the preceding equation was proven only for all F t -measurable functions φ1(μ), not for all F μ T -measurable functions. We rewrite the lemma as follows, stating equivalence between its two claims as well as a third and often more convenient form: Lemma 3.7*. Let P ∈ Pp(Ω) such that (B,W ) is a Wiener process with respect to the filtration (F t )t∈[0,T ] under P , and define ρ := P ◦ (ξ,B,W, μ)−1. Suppose that (1) and (3) of Definition 3.4 are satisfied, that P (X0 = ξ) = 1, and that the state equation (3.3) holds under P . The following are equivalent: (A) For P ◦ μ−1-almost every ν ∈ Pp(X ), it holds that (Wt)t∈[0,T ] is an (F W,Λ,X t )t∈[0,T ]-Wiener process under ν. (B) Under P , F T ∨ F ξ,W,Λ t is independent of σ{Ws −Wt : s ∈ [t, T ]} for every t ∈ [0, T ). (C) P is an MFG pre-solution Proof. (A⇒ C): Let Q = P ◦ (ξ,B,W, μ,Λ)−1. Assuming (A) holds, the second and third steps of the original proof [1, Lemma 3.7] are correct and show that Q ∈ A(ρ). As all of the other defining properties of an MFG pre-solution hold by assumption, we deduce (C). (C ⇒ B): Note that (C) entails that FΛ t is conditionally independent of F ξ,B,W,μ T given F ξ,B,W,μ t under P , for every t ∈ [0, T ). Fix t ∈ [0, T ), and fix bounded functions φt, ψT , ψt, and ht+ such that φt(Λ) is FΛ t -measurable, ψT (B,μ) is F B,μ T -measurable, ψt(ξ,W ) is F ξ,W t -measurable, and ht (W ) is σ{Ws −Wt : s ∈ [t, T ]}-measurable. The conditional independence yields E [ φt(Λ)| F T ] = E [ φt(Λ)| F t ] , a.s. The independence of ξ, (B,μ), and W easily implies that F T ∨F ξ,W t is independent of σ{Ws− Wt : s ∈ [t, T ]}, and we deduce E [φt(Λ)ψT (B,μ)ψt(ξ,W )ht+(W )] = E [ E [ φt(Λ)| F t ] ψT (B,μ)ψt(ξ,W )ht+(W ) ] = E [ E [ φt(Λ)| F t ] ψT (B,μ)ψt(ξ,W ) ] E [ht+(W )] = E [φt(Λ)ψT (B,μ)ψt(ξ,W )]E [ht+(W )] .\",\"PeriodicalId\":50763,\"journal\":{\"name\":\"Annals of Probability\",\"volume\":null,\"pages\":null},\"PeriodicalIF\":2.1000,\"publicationDate\":\"2020-09-01\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"Annals of Probability\",\"FirstCategoryId\":\"100\",\"ListUrlMain\":\"https://doi.org/10.1214/20-aop1432\",\"RegionNum\":1,\"RegionCategory\":\"数学\",\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"Q1\",\"JCRName\":\"STATISTICS & PROBABILITY\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"Annals of Probability","FirstCategoryId":"100","ListUrlMain":"https://doi.org/10.1214/20-aop1432","RegionNum":1,"RegionCategory":"数学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"STATISTICS & PROBABILITY","Score":null,"Total":0}
引用次数: 0

摘要

这个注释修正了我们文[1]中的引理3.7。论文的主要结果仍然是正确的。这个注释纠正了[1,引理3.7]中的一个错误。引理并不像所说的那样正确,第一个结论必须作为一个假设来陈述。这个勘误表纠正了引理的陈述,然后表明在本文后面的引理的三个应用中,附加假设都是满足的。论文的主要结果保持不变。[1,引理3.7]中的错误出现在证明的“第一步”的末尾。具体来说,“第二步”之前的最后一个方程(第3769页第5-6行)是不准确的,因为前面的方程只被证明适用于所有的Ft-可测量函数φ1(μ),而不是适用于所有Fμt-可测量的函数。我们将引理改写如下,陈述了它的两个声明之间的等价性,以及第三种通常更方便的形式:引理3.7*。设P∈Pp(Ω),使得(B,W)是关于P下的过滤(Ft)t∈[0,t]的Wiener过程,并定义ρ:=P◦ (ξ,B,W,μ)−1。假设满足定义3.4的(1)和(3),P(X0=ξ)=1,并且状态方程(3.3)在P下成立。以下是等效的:(A)对于P◦ μ−1-几乎每一个Γ∈Pp(X),它认为(Wt)t∈[0,t]是Γ下的(FW,∧,Xt)t∈[0],t]-Wiener过程。(B) 在P下,对于每一个T∈[0,T),F∈Fξ,W,∧T独立于σ{Ws−Wt:s∈[T,T]}。(C)P是MFG预解证明⇒ C) :设Q=P◦ (ξ,B,W,μ,∧)−1。假设(A)成立,则原证明[1,引理3.7]的第二步和第三步是正确的,并证明Q∈A(ρ)。由于MFG预解的所有其他定义性质都是通过假设成立的,我们推导出(C)。(C⇒ B) :注意(C)要求F∧t条件独立于Fξ,B,W,μt,给定Fξ、B,W、μt在P下,对于每个t∈[0,t).固定t∈[0],t),并固定有界函数φt,ψt,ψt和ht+,使得φt(∧)是F∧t-可测的,ψ(B,μ)是Fξ,t]}-可测量。条件独立性得到E[φt(∧)|F t]=E[φt(∧,并推导出E[φt(∧)ψt(B,μ)ψt(ξ,W)ht+。
本文章由计算机程序翻译,如有差异,请以英文原文为准。
Errata: Mean field games with common noise
This note corrects Lemma 3.7 in our paper [1]. The main results of the paper remain correct as stated. This note corrects an error in [1, Lemma 3.7]. The lemma is not correct as stated, and the first conclusion must instead be stated as a hypothesis. This erratum corrects the statement of the lemma and then shows that the additional hypothesis is satisfied in each of the three applications of the lemma later in the paper. The main results of the paper remain unchanged. The error in [1, Lemma 3.7] is at the end of the “first step” of the proof. Specifically, the last equation before the “second step” (lines 5-6 of page 3769) is not accurate, because the preceding equation was proven only for all F t -measurable functions φ1(μ), not for all F μ T -measurable functions. We rewrite the lemma as follows, stating equivalence between its two claims as well as a third and often more convenient form: Lemma 3.7*. Let P ∈ Pp(Ω) such that (B,W ) is a Wiener process with respect to the filtration (F t )t∈[0,T ] under P , and define ρ := P ◦ (ξ,B,W, μ)−1. Suppose that (1) and (3) of Definition 3.4 are satisfied, that P (X0 = ξ) = 1, and that the state equation (3.3) holds under P . The following are equivalent: (A) For P ◦ μ−1-almost every ν ∈ Pp(X ), it holds that (Wt)t∈[0,T ] is an (F W,Λ,X t )t∈[0,T ]-Wiener process under ν. (B) Under P , F T ∨ F ξ,W,Λ t is independent of σ{Ws −Wt : s ∈ [t, T ]} for every t ∈ [0, T ). (C) P is an MFG pre-solution Proof. (A⇒ C): Let Q = P ◦ (ξ,B,W, μ,Λ)−1. Assuming (A) holds, the second and third steps of the original proof [1, Lemma 3.7] are correct and show that Q ∈ A(ρ). As all of the other defining properties of an MFG pre-solution hold by assumption, we deduce (C). (C ⇒ B): Note that (C) entails that FΛ t is conditionally independent of F ξ,B,W,μ T given F ξ,B,W,μ t under P , for every t ∈ [0, T ). Fix t ∈ [0, T ), and fix bounded functions φt, ψT , ψt, and ht+ such that φt(Λ) is FΛ t -measurable, ψT (B,μ) is F B,μ T -measurable, ψt(ξ,W ) is F ξ,W t -measurable, and ht (W ) is σ{Ws −Wt : s ∈ [t, T ]}-measurable. The conditional independence yields E [ φt(Λ)| F T ] = E [ φt(Λ)| F t ] , a.s. The independence of ξ, (B,μ), and W easily implies that F T ∨F ξ,W t is independent of σ{Ws− Wt : s ∈ [t, T ]}, and we deduce E [φt(Λ)ψT (B,μ)ψt(ξ,W )ht+(W )] = E [ E [ φt(Λ)| F t ] ψT (B,μ)ψt(ξ,W )ht+(W ) ] = E [ E [ φt(Λ)| F t ] ψT (B,μ)ψt(ξ,W ) ] E [ht+(W )] = E [φt(Λ)ψT (B,μ)ψt(ξ,W )]E [ht+(W )] .
求助全文
通过发布文献求助,成功后即可免费获取论文全文。 去求助
来源期刊
Annals of Probability
Annals of Probability 数学-统计学与概率论
CiteScore
4.60
自引率
8.70%
发文量
61
审稿时长
6-12 weeks
期刊介绍: The Annals of Probability publishes research papers in modern probability theory, its relations to other areas of mathematics, and its applications in the physical and biological sciences. Emphasis is on importance, interest, and originality – formal novelty and correctness are not sufficient for publication. The Annals will also publish authoritative review papers and surveys of areas in vigorous development.
×
引用
GB/T 7714-2015
复制
MLA
复制
APA
复制
导出至
BibTeX EndNote RefMan NoteFirst NoteExpress
×
提示
您的信息不完整,为了账户安全,请先补充。
现在去补充
×
提示
您因"违规操作"
具体请查看互助需知
我知道了
×
提示
确定
请完成安全验证×
copy
已复制链接
快去分享给好友吧!
我知道了
右上角分享
点击右上角分享
0
联系我们:info@booksci.cn Book学术提供免费学术资源搜索服务,方便国内外学者检索中英文文献。致力于提供最便捷和优质的服务体验。 Copyright © 2023 布克学术 All rights reserved.
京ICP备2023020795号-1
ghs 京公网安备 11010802042870号
Book学术文献互助
Book学术文献互助群
群 号:481959085
Book学术官方微信