报告人：马俊杰

时    间：2025/01/17 13:00-17:00

地    点：线上报告

腾讯会议ID：321-587-300

摘    要：In the first part of this talk, we present a unified framework for constructing Approximate Message Passing (AMP) algorithms for rotationally-invariant models. By employing a general iterative algorithm template and reducing it to long-memory Orthogonal AMP (OAMP), we systematically derive the correct Onsager terms of AMP algorithms. This approach allows us to re-derive an AMP algorithm introduced by Fan and Opper et al., while shedding new light on the role of free cumulants of the spectral law. The free cumulants arise naturally from a recursive centering operation, potentially of independent interest beyond the scope of AMP.

In the second part of this talk, we consider the applications of our framework to signal estimation in spiked models with rotationally-invariant noise. We develop a new class of AMP algorithms and show that the resulting algorithm achieves the smallest possible asymptotic estimation error among a broad class of iterative algorithms under a fixed iteration budget.

This talk is based on joint work with Songbin Liu (AMSS, CAS) and Rishabh Dudeja (UW-Madison).

个人简介：马俊杰，中国科学院数学与系统科学研究院副研究员。2010年本科毕业于西安电子科技大学，2015年在香港城市大学取得博士学位。曾于香港城市大学、哥伦比亚大学和哈佛大学从事博士后研究。研究兴趣包括通信信号处理、随机矩阵、高维统计、机器学习等。曾入选中科院百人计划，主持自然基金青年项目并参与中科院先导科技专项、科技部重点专项等科研项目。目前担任中国运筹学会青年工作委员会副秘书长。

欢迎各位老师和同学参加！

联系人：刘伟华

邮    箱：lwh.math@zju.edu.cn