Regulated machine learning models in finance

发布时间:2022-12-16 浏览次数:10

报告人:Dangxing Chen, Assistant Professor, Duke Kunshan University

报告时间:  20221220日(周二)下午3:00-4:00

报告会场:腾旭会议 773-6902-9240

 

Abstract: For many years, machine learning methods have been used in a wide range of fields, including computer vision and natural language processing. Even though machine learning methods significantly improve model performance over traditional methods, their black-box structure makes it difficult for researchers to interpret the results. For highly regulated sectors, such as the financial sector, transparency, explainability, conceptual soundness, and fairness are at least as important as accuracy. In the absence of meeting regulatory requirements, even highly accurate machine learning methods will be unlikely to be accepted. To address the issue of transparency and interpretability, we introduce a novel class of machine learning algorithms known as generalized gloves of neural additive models. The generalized gloves of neural additive models separate features into three categories: linear features, individual nonlinear features, and interacted nonlinear features. Additionally, interactions in the last category are only local. In order to ensure conceptual soundness and fairness, we impose individual and pairwise monotonicity restrictions on the model, whereas the latter has been neglected in the literature to date. We demonstrate empirically that our models provide optimal accuracy with the simplest architecture with reasonable fairness, resulting in highly accurate and regulated machine learning.

Bio: Dangxing Chen is an Assistant Professor of Mathematics at Duke Kunshan University. He received his B.S. in Applied Mathematics from the University of Michigan at Ann Arbor in 2013 and his Ph.D. from the University of North Carolina at Chapel Hill in Applied Mathematics in 2017. His thesis work centered on developing fast and accurate numerical algorithms for partial differential equations and integral equations arisen in science and engineering with the emphasis on the large-scale long-time simulation. From 2017 to 2021, he held a postdoctoral position at the University of California, Berkeley. His research interests include machine learning methods in finance, quantitative finance, and numerical partial differential equations.


 
地址:苏州大学十梓街1号
版权所有 Copyright © 2012 苏州大学金融工程研究中心. 访问次数: