报告题目：Lp-norm based representation learning: some theories, algorithms, and applications [Slides]
 Xi Peng, Zhang Yi, and Huajin Tang, Robust Subspace Clustering via Thresholding Ridge Regression,
The Twenty-Ninth AAAI Conference on Artificial Intelligence (AAAI), Austin, Texas, USA, January 25–29, 2015;
 Xi Peng, Zhiding Yu, Huajin Tang, and Zhang Yi, Constructing L2-Graph for Subspace Learning and Segmentation, arXiv1209.0841;
 Xi Peng, Canyi Lu, Zhang Yi, and Huajin Tang, Connections Between Nuclear Norm and Frobenius Norm Based Representation, arXiv1502.07423;
 Xi Peng, Jiwen Lu, Yan Rui, and Zhang Yi, Automatic Subspace Learning via Principal Coefficients Embedding, arXiv1411.4419;
报告摘要：In this talk, I will introduce some works on Lp-norm based representation learning. The talk consists of three parts. First, I will introduce a Frobenius-norm based representation learning method and its applications in subspace clustering and subspace learning. Second, I will introduce a theoretical study on the connections between Frobenius norm based representation and nuclear-norm based representation. Finally, I will introduce an automatic subspace learning method that can automatically estimate the feature dimension and achieve robust results from corrupted data.
报告人简介：彭玺，目前是新加坡（Institute for Infocomm., Research Agency for Science, Technology and Research (A*STAR)）信息通信研究所研究员(Research Scientist)。他于2013年在四川大学计算机学院章毅教授指导下获得博士学位。主要研究兴趣是无监督的表达学习（unsupervised representation learning）及其在计算机视觉和机器学习中的理论、算法及应用，目前在CVPR，AAAI，IJCAI，TNNLS，TCYB等国际会议及期刊上发表论文多篇，是多个国际会议及期刊例如AAAI，TNNLS，TKDE的审稿人。
报告题目：Transitive Distance Clustering: Theories, Algorithms and Applications [Slides]
 Zhiding Yu, Weiyang Liu, Wenbo Liu, Yingzhen Yang, Ming Li and B. V. K. Vijaya Kumar. On Order-Constrained Transitive Distance Clustering. (AAAI 2016)
 Zhiding Yu, Weiyang Liu, Wenbo Liu, Xi Peng, Zhuo Hui and B. V. K. Vijaya Kumar, Generalized Transitive Distance with Minimum Spanning Random Forest. (IJCAI 2015)
 Zhiding Yu, Chunjing Xu and Deyu Meng et al. Transitive Distance Clustering with K-Means Duality. (CVPR 2014)
报告摘要：Transitive distance (TD) is an ultrametric with elegant properties for data clustering. Suppose a path is composed by a sequence of data points and edges and a “gap” is the largest edge along this path. Then given any pairwise data, their transitive distance is defined as the smallest possible “gap” along all the paths that connects them. Such metric definition renders TD capability of addressing highly elongated and non-convex cluster structures by significantly reducing the intra-cluster distances. In this talk, I will introduce the concept of TD, its properties and theories, as well as several latest improvements on top of the conventional method. I will also talk about multiple applications of the proposed clustering methods, with state of the art performance in image clustering, speech data clustering and image segmentation.
报告人简介：Zhiding Yu is currently a 4th year Ph.D. candidate with the Department of Electrical and Computer Engineering, Carnegie Mellon University. He graduated with a B.Eng. degree from the Elite Class of Electrical Engineering, South China University of Technology in 2008, and obtained the M.Phil. degree from the Department of Electronic and Computer Engineering, Hong Kong University of Science and Technology in 2012. His main research interests include structured prediction for scene understanding, object detection, clustering and image segmentation. He was twice the recipient of the HKTIIT Post-Graduate Excellence Scholarships (2010/2012). He is a co-author of the best student paper in International Symposium on Chinese Spoken Language Processing (ISCSLP) 2014, and the winner of best paper award in IEEE Winter Conference on Applications of Computer Vision (WACV) 2015. He did internships at Adobe Research and Microsoft Research respectively in 2013 and 2015. His intern work on facial expression recognition at Microsoft Research won the First Runner Up at the EmotiW-SFEW Challenge 2015 and was integrated to the Microsoft Emotion Recognition API under Project Oxford.