4月4日下午，应伟德betvlctor1946和科技处邀请，美国哈佛大学计算机科学与生物信息学研究员王海帅做了题目为《针对大规模复杂数据环境的高级数据分析》(Advanced Data Analytics for Large-scale Complex Data Environments)的学术报告。报告会在长清湖校区信息学院实验楼526会议室举行，由学院副院长郑元杰教授主持，学院部分老师、博士、硕士研究生聆听了此次报告。

报告会上，王海帅研究员首先从复杂数据对象的类型和传统的数据分析方法入手，引出了当前数据分析与研究的目标，即从复杂的数据环境中找出一些区分性特征用于具体任务的预测与分类。针对这类问题，他从定义、算法等方面向我们介绍了传统图结构(Graph Data Structure)、时间序列数据结构(Time Series Data Structure)、时间变量图结构(Time-Variant Graph Data Structure)和网络时间序列结构(Networked Time Series Data Structure)四种数据结构，并详细介绍了每一种数据结构在社交网络和生物信息或医疗健康领域中的研究与应用。随后王海帅研究员借助其参与的切身实例，具体分析阐述了复杂环境中数据分析的挑战问题，并对每一类问题具体介绍了解决方法。此外，他还就深度学习与传统方法在特征提取方面的特点做了具体的分析与对比。最后，王海帅研究员向在场师生展示了当前自己的研究成果和项目情况。

报告会问答环节，王海帅研究员和在座师生就报告内容展开了激烈的讨论，并对现场老师、学生提出的有关子图挖掘、原始数据采集与处理、深度学习等研究领域的相关问题做了认真解答。报告会气氛活跃，内容清晰详实，使我院师生对大数据分析的方法和应用有了更深的理解与知识，对今后的科学研究有着重要的指导意义。

Haishuai Wang is a Research Fellow in Computer Science and Bioinformatics at Harvard University. Before Harvard, he was a Postdoctoral Associate in the Department of Computer Science at Washington University in St. Louis. Dr. Wang received his Ph.D. degree in Computer Science from the Center of Artificial Intelligence (CAI) at the University of Technology Sydney. His research focuses on machine learning, data mining and deep learning. He has been applying his research to real-world applications, including large-scale optimization for social network analysis and computational approaches to genomic and clinical data. Dr. Wang has a long-term research interest in improving clinical outcomes using artificial intelligence algorithms on high-dimensional genomic data and electronic medical records. Dr. Wang serves as Program Committee (PC) for top conferences (e.g., KDD, NIPS and SDM), and Associate Editor for IEEE Access and Journal of Health & medical informatics. He also has been with a number of companies for internships, including IBM, Hitachi and Intel. His work is supported by Google, Amazon, NIH, Harvard Data Science etc.