报告题目：Economics-Aware Gaussian Process Modelling for Option-Implied Risk Metrics

报告人： 陈泽汛 博士 （英国爱丁堡大学）

报告时间：2024年12月05日下午 19:00-20:00

报告地点：腾讯会议（693-776-383）

报告摘要：Machine learning models are predominantly data-driven and often lack embedded domain knowledge. This limitation is particularly significant in the field of finance, where certain asset conditions must be maintained. To address this, we propose a novel constrained Gaussian Process model (consGP) that simultaneously minimises interpolation loss and satisfies encoded linear inequalities representing economic constraints. This approach enables the consGP to learn from market data whilst adhering to fundamental economic principles. We apply this model to the estimation of option-implied risk metrics, where the consGP demonstrates robust performance in estimating risk-neutral density (RND) across sparse and noisy option observations. This model has been demonstrated to be particularly suitable for modeling stock options with limited sample sizes due to insufficient liquidity. Our comprehensive empirical studies, conducted using a cross-section of S\&P 500 stocks, reveal that the consGP model outperforms traditional structural models in recovering stock-level RND. This improved performance translates into enhanced predictive information and tangible economic benefits for investors. The consGP model thus represents a significant advancement in integrating machine learning techniques with domain-specific financial constraints, offering a more robust and economics-aware approach to option pricing and risk assessment.

报告人简介：陈泽汛，英国爱丁堡大学商学院助理教授，研究方向包括非参贝叶斯统计，概率化统计学习方法，时间序列分析，算法公平性与信用评级，多维度人口流动模型，机器学习方法在人口流动与城市规划上的应用，高斯过程机器学习在金融期权与市场风险上的应用等等。陈博士2013年本科毕业于山东大学数学与应用数学专业，2017年博士毕业于英国莱斯特大学数学系，研究方向是高斯过程机器学习算法及其推广与应用。博士毕业后曾在英国苏塞克斯大学作为博士后研究员参与英国EPSRC项目，提出高斯过程分类器的公平性改进。之后又在英国埃克塞特大学作为博士后参与英美联合US Army Research的项目，研究内容包括时间，空间与社交数据下人口流动问题的建模等。陈博士曾在各类期刊会议上发表过一系列高水平文章，其中包括Nature Communications，Neural Computation and Application，Measurement等等，目前是英国皇家统计协会会员, 英国皇家统计协会会员青年统计学家分会组委会秘书，英国数学与应该数学协会会员，国际统计协会会员，并担任各类数学统计，量化金融，机器学习，工程机械，人口流动，交通运输方面期刊的审稿人，如IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI)，Association for the Advancement of Artificial Intelligence (AAAI)，Transportation Research Part C: Emerging Technologies，The European Journal of Finance， Journal of Forecasting等。