时间：2018年12月26（星期三）14：00-15：00

地点：学院南路校区，学术会堂603

报告题目：RecentResults of Data Mining from High-Dimensional Data

报告人：Qiang Cheng, Department of Internal Medicine, University ofKentucky

报告摘要：Diverseareas of scientific research and everyday life are now deluged withhigh-dimensional data. There is a pending need of techniques for findingpatterns and discovering knowledge from them. In this talk I will present somerecent results of mining structural information or making predictions with suchdata, including unsupervised learning without labeling information, deepsupervised learning for detecting disease from medical time series data, anddeep supervised learning from network data. In particular, I will talk aboutrobust principal component analysis accounting for low-rank structures, aclustering model exploiting representation and learning jointly, and anadaptive graph classification model. Potential applications of them inbiomedicine, social science, and other areas are expected.

报告人简介：Dr. Cheng is Associate Professor at Division ofBiomedical Informatics, Department of Internal Medicine and AssociateProfessor, Department of Computer Science of University of Kentucky. He obtained PhD in Electrical andComputer Engineering at University of Illinois at Urbana-Champaign; Urbana, MSin Applied Mathematics and Computer Science in Peking University.Dr. Cheng’s researchinterests are in data science, machine learning, pattern recognition,signal/image processing, and biomedical/healthcare informatics. He isdeveloping data mining and machine learning techniques to help advancebiomedical research, for example, to integrate medical data from heterogeneoussources to find useful patterns, enable actionable insights, and facilitate data-drivenknowledge discovery. He handles diverse kinds of data such as small-sampledata, large-scale or big data, imaging data of various modalities, data withmixed types, and data with significant missing values. His work has appeared invarious venues such as IEEE TPAMI, TNNLS, TSP, CVPR, ICDM, AAAI, ICDE, ACMTIST, KDD, TKDD, CIKM, NIPS, and AAAI. He also has a number of patentsgenerated from his work.

本次活动受十大澳门信誉老牌网赌2018专题学术讲座项目资助。