专家姓名：Kit Ian KOU（高洁欣）

专家单位：澳门大学

联系方式：kikou@um.edu.mo

报告时间：2025年3月20日 10:00

报告地点： 格物楼3306室

报告题目：Robust Signal Recovery via Prolate Spherical Wave Functions and Maximum Correntropy Criterion

摘 要：Tis talk presents a novel approach for robust signal recovery by integrating prolate spherical wave functions (PSWFs) with the maximum correntropy criterion (MCC). Prolate spherical wave functions are renowned for their exceptional concentration properties in both time and frequency domains, making them highly effective for signal representation and recovery tasks. However, traditional methods based on PSWFs often struggle in the presence of noise or outliers. To address this limitation, we incorporate the maximum correntropy criterion, a powerful statistical measure that enhances robustness against non-Gaussian noise and outliers by maximizing the similarity between the estimated and true signals in a kernel-induced space.

Our proposed method combines the strengths of PSWFs and MCC to achieve superior signal recovery performance. Specifically, PSWFs provide an optimal basis for signal representation, while MCC ensures robustness against noise and outliers. We formulate the signal recovery problem as an optimization task and develop an efficient algorithm to solve it. Extensive experiments demonstrate that our approach significantly outperforms existing methods in terms of recovery accuracy and robustness, especially under challenging noise conditions.

专家简介：Kit Ian KOU（高洁欣），澳门大学数学系副教授、博士生导师。高老师在高维信号分析理论及其应用领域拥有20余年的研究经验，尤其专注于四元数、傅里叶变换和张量分析在高维信号处理中的应用。主持或作为核心成员参与科研项目27项，项目资助方包括澳门科技发展基金、国家自然科学基金、广东省科技局等。发表高水平学术论文130余篇，其代表性成果之一《四元数和线性正则分析在彩色人脸识别和边缘检测中的应用》荣获2018年度澳门科学技术奖（自然科学奖）。高老师曾担任多个国际期刊的特邀编辑和审稿人。目前，她是英国剑桥大学Clare Hall的终身会员，并任职于澳门应用数学研究中心。

编辑：缪吉飞  责任编辑：鲁学伟  审核：徐瑾