Biosketch: Prof. Juntao Gao is the Research Associate Professor, Beijing National Research Center for Information Science and Technology (BNRist), Tsinghua University. He received Ph.D degree from Heidelberg University, Germany in 2005 (supervisor: Prof. Roland Eils) and received postdoctoral training in Stowers Institute for Medical Research in USA (supervisor: Prof. Rong Li). He has been working in Tsinghua University since 201, being engaged in research on developing imaging and bioinformatic methods to study 3D genome and gene regulation. He has published more than 50 papers in PNAS, Nature Communications, Nucleic Acid Research, Light: Science & Applications (LSA), IEEE Transactions on Medical Imaging, Redox Biology, ACS Photonics, and other journals. Some work has been reported by Nature Methods as "Research Highlights‟.
Speech Title: Big data in image-based spatial multi-omic database iSMOD
Abstract: The very first integrative browser iSMOD (imaged-based Spatial Multi-omic Database), which includes imaged-based spatial genomic, spatial transcriptomic, and the spatial information of key nucleic proteins in cell nucleus, was developed by us recently(https://www.i-smod.com/). iSMOD is constructed to collect and browse the comprehensive FISH and nucleus proteomics information from the title, abstract, keywords, and all relevant images of 40,000+ (and still growing) public papers in Pubmed. An automatic figure parser method was developed here, including high precision subfigure segmentation, screening, figure-text extraction, caption matching and reference extraction, etc., to obtain all essential information.
As the very first browser for imaged-based spatial multi-omic data focusing on the key players in cell nucleus, iSMOD provides a corner stone to browse the images of different genomic loci and key proteins, to verify various chromatin interactions in different species, and to build up a 3D model of chromatin in cell nuclei of different species, in order to serve the field and scientific community in a more efficient way.
Biosketch: Dr. Fei Hao received the Ph.D. degree in Computer Science and Engineering from Soonchunhyang University, South Korea, in 2016. Since 2016, he has been with Shaanxi Normal University, Xi'an, China, where he is an Associate Professor. From 2020 to 2022, he was a Marie Sklodowska-Curie Fellow with the University of Exeter, Exeter, United Kingdom. His research interests include social computing, ubiquitous computing, big data analytics, knowledge graph and edge intelligence. He is also a China Regional Director of International Association for Convergence Science and Technology (IACST), an executive director of Shanxi Association of Experts and Scholars (SAES) Information Branch, and an executive director of Shanxi Block-chain Research Association. Dr. Hao holds a world-class research track record of publication in the top international journals and the prestigious conferences. He has published more than 150 papers in the leading international journals and conference proceedings, such as IEEE Transactions on Parallel and Distributed Systems, IEEE Transactions on Services Computing, IEEE Communications Magazine, IEEE Internet Computing, ACM Transactions on Multimedia Computing, Communications and Applications as well as ACM SIGIR and GlobeCom. In addition, he was the recipient of 6 Best paper awards from CSA 2020, CUTE 2016, UCAWSN 2015, MUE2015, IEEE GreenCom 2013 and KISM 2012 conferences, respectively. He was also the recipient of the Outstanding Service Award at SMMA 2020, FutureTech 2019, DSS2018, and SmartData 2017, the IEEE Outstanding Leadership Award at IEEE CPSCom 2013 and the 2015 Chinese Government Award for Outstanding Self-Financed Students Abroad. Since 2017, he has joined JIPS (Journal of Information Processing Systems) editorial board, where he is currently an associate editor. He is currently an editor of ICT Express journal. And he is an initiator and general/program chair of IEEE DSCI and SMMA. He is also a member of ACM, CCF and KIPS.
Speech Title: Socially-aware Dependent Tasks Offloading in Mobile Edge Computing
Abstract: In recent years, with the advent of the 5G era and the continuous development of science and technology, the number of computation-intensive applications has been increasing. Faced with the massive amount of data generated by these computation-intensive applications and the lack of computing resources due to the hardware constraints of users' mobile devices, there is an inability to execute these applications. Mobile Edge Computing (MEC) speeds up the task transmission time by offloading the tasks to the edge server for execution, reducing the user's waiting time and improving the quality of service. However, when the computation-intensive applications are offloaded, these applications are usually decomposed into multiple tasks with dependencies, and the offloading of these dependent tasks needs to be offloaded in strict execution order, thus a huge challenge is raised for task offloading and its optimization. Moreover, in the case of offloading tasks with other devices, it is often required to consider the success rate of offloading , since not all users are willing to lend their mobile devices to others for task execution. To conquer this challenge, by taking social relationships between users into account, this research intends to combine the computational resources of local devices and edge clouds and provide more flexible offloading and execution solutions, for achieving the efficient offloading of dependent tasks with the joint consideration of network latency and energy consumption. Technically, this talk will introduce a joint optimization of latency and reward dependent task offloading method, an evolutionary algorithm-based cloud-edge-end collaborative dependent tasks offloading method, and a socially-aware dependent tasks offloading method.
Biosketch: Rui Yang (Member, IEEE) received the B.Eng. degree in computer engineering and the Ph.D. degree in electrical and computer engineering from the National University of Singapore, Singapore, in 2008 and 2013, respectively. He is currently an Associate Professor with the School of Advanced Technology, Xi’an Jiaotong–Liverpool University, Suzhou, China, and an Honorary Lecturer with the Department of Computer Science, University of Liverpool, Liverpool, U.K. His research interests include machine learning based data analysis and applications. He has authored or coauthored more than 80 technical papers. Dr. Yang has also been a very active reviewer for many international journals and conferences. He is currently serving as associate editors for Neurocomputing and International Journal of Network Dynamics and Intelligence.
Speech Title: MSDAN: A Multi-Subdomain Adaptation Network for Single-Source to Single-Target Cross-Subject Motor Imagery Classification
Abstract: In the electroencephalography (EEG) based cross-subject motor imagery (MI) classification task, the device and subject problems can cause the time-related data distribution shift problem. In a single-source to single-target (STS) MI classification task, such a shift problem will certainly provoke an increase in the overall data distribution difference between the source and target domains, giving rise to poor classification accuracy. In this paper, a novel multi-subdomain adaptation method (MSDAN) is proposed to solve the shift problem and improve the classification accuracy of the traditional approaches. In the proposed MSDAN, the adaptation losses in both class-related and time-related subdomains (that are divided by different data labels and session labels) are obtained by measuring the distribution differences between the source and target subdomains. Then, the adaptation and classification losses in the loss function of MSDAN are minimized concurrently. To illustrate the application value of the proposed method, our method is applied to solve the STS MI classification task about data analysis with respect to the brain-computer interface (BCI) competition III-IVa dataset. The resultant experiment results demonstrate that compared with other well-known domain adaptation and deep learning methods, the proposed method is capable of solving the time-related data distribution problem at higher classification accuracy.