Tag
本文提出 BeTINH 结构化条件神经求解器，通过多方向时间-参数调制与联合残差路径提升 Heston 随机波动率 PDE 参数化族的求解精度与导数保真度，强调下游敏感度指标需导数感知的评估。
Proposes an uncertainty-aware reinforcement learning framework for algorithmic trading that integrates distributional, epistemic, and aleatoric uncertainty using SHAP-weighted reconstruction, MC Dropout, and LSTM consensus. Outperforms traditional models on five major US stock indices.