Skip to content


Co event with Department Colloquium

Hashing for Semantic Large-Scale Image Retrieval in Non-Stationary Environments

Dr. Wing W. Y. Ng
School of Computer Science and Engineering
South China University of Technology

Time: August 19th, 1:00pm to 3:00 pm

Location: Conference room 755 of Computer Science Department

A very large volume of images is uploaded to the Internet daily. Hashing methods provide efficient image retrieval results with sublinear time complexity. However, current hashing methods for image retrieval are designed for static databases only. They fail to consider the fact that the distribution of images can change when new images are added to the database over time. Changes in the distribution include both the discovery of new classes and changes to the distribution of images within a class owing to concept drifting. In this seminar, a brief introduction to hashing methods for large-scale image retrieval will be given. Then, new research results on hashing for non-stationary environments will be explored.

About the Speaker: Dr. Wing W. Y. Ng received his B.Sc. and Ph.D. degrees from Hong Kong Polytechnic University in 2001 and 2006, respectively. He is now a Professor with the School of Computer Science and Engineering, South China University of Technology, China. His major research interests are machine learning and image retrieval in big data and non-stationary environments. Prof. Ng has published 20+ journal papers in IEEE Transactions on Neural Networks, IEEE Transactions on Cybernetics, IEEE Transactions on Multimedia, etc. His works have been cited 1000+ times and his h-index is 16 (Google Scholar). His IEEE TNN publication proposing the localized generalization error model for neural networks has been cited 100+ times. He is the principal investigator for three China National Natural Science Foundation projects and the Program for New Century Excellent Talents in University from the China Ministry of Education. Prof. Ng is currently an associate editor of the International Journal of Machine Learning and Cybernetics. He is an IEEE senior member and served on the Board of Governors of the IEEE Systems, Man and Cybernetics Society (SMCS) during 2011–2013.

Co event with Department Colloquium


Biological Applications of High-Performance Computing in Metagenomics and Metaproteomics

Dr. Chongle Pan

Time: Friday, Nov. 13rd, 2015

Location: Conference Room #755

Human microbiomes and environmental microbial communities drive many important biological processes in their ecosystems. The taxonomic composition and genetic potential of a microbial community can be measured by metagenomics. We developed the Sigma algorithm to identify different strains of microorganisms in a microbial community and find their genomic variations from metagenomic sequencing. A nonlinear programming method was used to solve a maximum likelihood estimation problem in the Sigma probabilistic model. This allowed rigorous assignment of ambiguous reads that can be aligned to multiple genomes. Sigma can be used for metagenomic biosurveillance as demonstrated by the identification of a pathogenic E. coli strain from a fecal sample. We addressed the computational challenge of PTM identification in metaproteomes by using a scalable database searching algorithm, Sipros, and the Titan supercomputer. We found extensive PTMs on enzymes involved in central metabolisms, chemotaxis, and other cellular functions in a model microbial community and uncovered dynamic regulation of these PTMs between two growth stages of this community.

About the Speaker: Dr. Chongle Pan is a staff scientist in the Computer Science and Mathematics Division of Oak Ridge National Laboratory and a joint associate professor in the Department of Microbiology of the University of Tennessee. He is interested in the characterization of microbial communities using integrated omics analyses. His computational research is focused on analysis of big data from metagenomics and metaproteomics measurements with scalable distributed computing. Omics results provide the foundation for studying microbial ecology and evolutionary biology in ecosystems of relevance to global carbon cycling and human health.    



Co Event with Department Colloquium

Big Graph Processing: Parallel Abstractions and Optimizations 

Dr. Ling Liu

Time: September 28, 1:00pm to 2:00 pm

Location: Conference room 755 of Computer Science Department

Large-scale real-world graphs are known to have highly skewed vertex degree distribution and highly skewed edge weight distribution. Existing vertex-centric iterative graph computation models suffer from poor scalability due to a number of serious problems: poor load balance of parallel execution, inefficient CPU resource utilization with respect to the cost of in-memory or on-disk graph access, and insufficient optimizations for computational performance. In this talk, I will first discuss the main challenges for big graph processing and then present GraphTwist, a scalable, efficient, and provably correct two-tier graph parallel processing system. At the storage and access tiers, GraphTwist maximizes parallel efficiency by employing three graph parallel abstractions for partitioning a big graph by slice-, strip-, or dice-based partitioning techniques. At the computation tier, GraphTwist presents two utility-aware pruning strategies: slice pruning and cut pruning, to further improve computational performance while preserving the computational utility defined by graph applications. I will end the talk by presenting some interesting research problems and unique opportunities in high-performance large-scale graph processing.

About the Speaker: Dr. Ling Liu is a professor in the School of Computer Science at the Georgia Institute of Technology. She directs the research programs in the Distributed Data Intensive Systems Lab (DiSL), examining various aspects of large-scale data-intensive systems, including performance, availability, security, and privacy. Prof. Liu is an elected IEEE Fellow and a recipient of the IEEE Computer Society Technical Achievement Award in 2012. She has published over 300 international journal and conference articles and is a recipient of the best paper award from a number of top venues, including ICDCS 2003, WWW 2004, 2005 Pat Goldberg Memorial Best Paper Award, IEEE Cloud 2012, IEEE ICWS 2013, Mobiquitous 2014, and ACM/IEEE CCGrid 2015. In addition to serving as general chair and PC chair for numerous IEEE and ACM conferences in the data engineering, very large databases, distributed computing, and cloud computing fields, Prof. Liu has served on the editorial board of over a dozen international journals. Currently, Prof. Liu is the editor-in-chief of IEEE Transactions on Service Computing. Prof. Liu’s current research is primarily sponsored by NSF, IBM, and Intel.