arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2604.13147stat.MLcs.LGmath.PR
收藏

基于离模型训练和重要性采样的自适应学习用于完全非马尔可夫最优随机控制(完整版)

Adaptive Learning via Off-Model Training and Importance Sampling for Fully Non-Markovian Optimal Stochastic Control. Complete version

Departamento de Matemática, Universidade de Brasília(数学系,巴西利亚大学);Università di Pisa, DEM(比萨大学,DEM);Université Paris Cité, LPSM(巴黎Cité大学,LPSM)

Dorival Leão, Alberto Ohashi, Simone Scotti, Adolfo M. D da Silva

AI总结:

针对完全非马尔可夫且依赖未知模型参数的连续时间随机控制问题,提出一种基于离散骨架和重要性采样的蒙特卡洛学习方法,实现离模型训练架构和自适应参数更新,并给出非渐近误差界。

摘要:

本文研究连续时间随机控制问题,其受控状态是完全非马尔可夫的,且依赖于未知模型参数。这类问题自然出现在路径依赖随机微分方程、粗糙波动率对冲以及分数布朗运动驱动的系统中。基于先前工作中发展的离散骨架方法,我们提出了一种用于相关嵌入后向动态规划方程的蒙特卡洛学习方法。我们的主要贡献有两方面。首先,针对几类具有代表性的非马尔可夫受控系统,我们构造了显式的支配训练律和Radon-Nikodym权重。这产生了一种离模型训练架构,其中在参考律下生成固定的合成数据集,而通过重要性采样恢复与目标模型相关的动态规划算子。其次,我们利用这种结构设计了参数模型不确定性下的自适应更新机制,使得可以通过重新加权相同的训练样本而非重新生成新轨迹来执行重复校准。对于固定参数,我们建立了通过深度神经网络逼近嵌入动态规划方程的非渐近误差界。对于自适应学习,我们推导了将蒙特卡洛逼近误差与模型风险误差分离的定量估计。数值实验在结构化线性二次型例子中展示了离模型训练机制和自适应重要性采样更新。

英文摘要:

This paper studies continuous-time stochastic control problems whose controlled states are fully non-Markovian and depend on unknown model parameters. Such problems arise naturally in path-dependent stochastic differential equations, rough-volatility hedging, and systems driven by fractional Brownian motion. Building on the discrete skeleton approach developed in earlier work, we propose a Monte Carlo learning methodology for the associated embedded backward dynamic programming equation. Our main contribution is twofold. First, we construct explicit dominating training laws and Radon--Nikodym weights for several representative classes of non-Markovian controlled systems. This yields an off-model training architecture in which a fixed synthetic dataset is generated under a reference law, while the dynamic programming operators associated with a target model are recovered by importance sampling. Second, we use this structure to design an adaptive update mechanism under parametric model uncertainty, so that repeated recalibration can be performed by reweighting the same training sample rather than regenerating new trajectories. For fixed parameters, we establish non-asymptotic error bounds for the approximation of the embedded dynamic programming equation via deep neural networks. For adaptive learning, we derive quantitative estimates that separate Monte Carlo approximation error from model-risk error. Numerical experiments illustrate both the off-model training mechanism and the adaptive importance-sampling update in structured linear-quadratic examples.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑