深度神经网络优化Heston定价模型  

Deep Neural Network Optimization of Heston Pricing Model

作　　者：闫海波 张馨月 YAN Haibo;ZHANG Xinyue(College of Statistics and Data Science,Xinjiang University of Finance and Economics,Urumqi,China,830012)

机构地区：[1]新疆财经大学统计与数据科学学院,乌鲁木齐830012

出　　处：《福建电脑》2025年第4期9-14,共6页Journal of Fujian Computer

摘　　要：为提升Heston定价模型在金融衍生品定价中的准确性,本文构建一个预测期权收盘价的PINN_Heston模型。采用结合PINN与Heston方程,构造物理方程约束训练的神经网络。模型以标的资产价格、到期时间及随机波动率为输入,期权价格为输出。实验的结果显示,PINN_Heston方法在定价上优于传统Heston模型和Black-Scholes模型,说明深度学习方法在期权定价领域具有优势,可为金融模型的改进和创新提供一个新思路。To improve the accuracy of Heston pricing model in financial derivatives pricing,this paper constructs a PINN_Heston model for predicting the closing price of options.Construct a neural network trained with physical equation constraints by combining PINN and Heston equations.The model takes the underlying asset price,expiration time,and random volatility as inputs,and solves the option price as the output.The experimental results show that the PINN_Heston method outperforms traditional Heston models and Black Scholes models in pricing,indicating that deep learning methods have advantages in the field of option pricing and can provide a new idea for the improvement and innovation of financial models.

关 键 词：神经网络 Heston模型 BLACK-SCHOLES模型 期权定价 

分 类 号：F224.0[经济管理—国民经济]