On the Martingale Problem for Measure-Valued Markov Branching Processes

On the Martingale Problem for Measure-Valued Markov Branching Processes
复制标题

DOI:
10.1007/978-1-4612-0381-0_4
复制
发表时间:
1992
期刊:
--
影响因子:
--
通讯作者:
P. Fitzsimmons
P. Fitzsimmons
中科院分区:
其他
文献类型:
--
作者:
P. Fitzsimmons

文献摘要

被引文献

相似文献

用[F]中的定理(4.1)(b)说明了与(ξ, ø)- superprocessx相关的鞅问题具有唯一解。定理是正确的,但[F]中提供的证明是不充分的,因为p. 355上t→Vtf(x)连续可微的断言一般是错误的。我们在这篇笔记中的目的是给出这个结果的一个完整的证明。由于在[F]中使用的假设的弱点,证明相当复杂,但由于该定理最近得到了应用(例如[DP]),我们认为有必要进行彻底的处理。此外,关于时空鞅的主要辅助结果(定理(2.22))可能具有独立的兴趣。在本节的剩余部分,我们回顾[F]中的一些符号,并陈述正在讨论的定理。所有未解释的符号如[F]所示。
It was stated as Theorem (4.1)(b) in [F] that the martingale problem associated with the (ξ, ø)-superprocessXdiscussed in that paper has a unique solution. The theorem is true, but the proof supplied in [F] is inadequate since the assertion on p. 355 thatt→Vtf(x) is continuously differentiable is false in general. Our purpose in this note is to give a complete proof of this result. Owing to the weakness of the hypotheses employed in [F], the proof is rather involved, but since the theorem has found recent application (e.g. [DP]), we thought a thorough treatment was warranted. Also, the main auxiliary result (Theorem (2.22)) concerning space-time martingales may be of independent interest. In the remainder of this section we recall some notation from [F] and state the theorem under discussion. All unexplained notation is as in [F].