主讲人：欧阳乐 深圳大学副研究员

时间：2019年10月28日16：00

地点：三号楼301报告厅

举办单位：数理学院

主讲人介绍：欧阳乐，博士，副研究员，硕士研究生导师，IEEE会员。于2015年6月获中山大学概率论与数理统计专业统计模式识别方向理学博士学位；2013年10月至2014年4月，赴新加坡南洋理工大学计算机科学系交流访问。2015年-2016年在香港城市大学电子工程系从事博士后研究。主要从事数据挖掘、机器学习和生物信息学等领域的科研和教学工作。目前已在IEEE  TCYB，Bioinformatics，BMC Bioinformatics，BMC Genomics, IEEE/ACM TCBB, Methods,  Scientific Reports，Molecular  BioSystems等国际权威期刊发表SCI论文30篇。获授权专利1项。担任Bioinformatics、PLoS Computational  Biology、Methods、IEEE/ACM TCBB等重要刊物审稿人。主持国家自然科学基金等项目。

内容介绍：Graphical models have been widely used to learn the conditional dependence  structures among random variables. In many controlled experiments, such as the  studies of disease or drug effectiveness, learning the structural changes of  graphical models under two different conditions is of great importance. However,  most existing graphical models are developed for estimating a single graph and  based on a tacit assumption that there is no missing relevant variables, which  wastes the common information provided by multiple heterogeneous data sets and  underestimates the influence of latent/unobserved relevant variables. In this  work, we propose a joint differential network analysis (JDNA) model to jointly  estimate multiple differential networks with latent variables from multiple data  sets. Extensive simulation experiments demonstrate that JDNA model outperforms  state-of-the-art methods in estimating the structural changes of graphical  models. Moreover, a series of experiments on several real-world data sets have  been performed and experiment results consistently show that our proposed JDNA  model is effective in identifying differential networks under different  conditions.