听众报名 | 第四届青年论坛日程和报告人来啦

关键词：招聘  青年论坛

北京大学前沿计算研究中心第四届青年论坛将于2019年10月23日在北京大学静园五院举办，旨在为计算理论、人工智能等领域以及相关交叉领域的海内外青年学者提供高水平学术交流平台，吸引国际优秀人才加盟中心。同时，中心将邀请目前活跃在工业界的技术精英和领袖，分享当前国内相关领域最新技术。我们诚邀海内外优秀学者和校友相聚燕园，交流学术前沿热点，探讨未来科技发展。

论坛日程

特邀报告人

叶杰平 Jieping Ye

Head of Didi AI Labs & VP of Didi Chuxing

Professor, University of Michigan

Biography

Dr. Jieping Ye is head of Didi AI Labs and a VP of Didi Chuxing. He is also a professor of University of Michigan, Ann Arbor. His research interests include big data, machine learning, and data mining with applications in transportation and biomedicine. He has served as a Senior Program Committee/Area Chair/Program Committee Vice Chair of many conferences including NeurIPS, ICML, KDD, AAAI, IJCAI, CIKM, ICDM, and SDM. He has served as an Associate Editor of Data Mining and Knowledge Discovery, IEEE Transactions on Knowledge and Data Engineering, and IEEE Transactions on Pattern Analysis and Machine Intelligence. He won the NSF CAREER Award in 2010. His papers have been selected for the outstanding student paper at ICML in 2004, the KDD best research paper runner up in 2013, and the KDD best student paper award in 2014.

AI for Transportation

Didi Chuxing is the world's leading mobile transportation platform that offers a full range of app-based transportation options for 550 million users. Every day, DiDi's platform receives over 100TB new data, processes more than 40 billion routing requests, and acquires over 15 billion location points. In this talk, I will show how AI technologies have been applied to analyze such big transportation data to improve the travel experience for millions of users.

熊辉 Hui Xiong

Head of Business Intelligence Lab & Talent Intelligence Center at Baidu

Professor, Rutgers University

Biography

Hui Xiong is a Professor at Rutgers University, and is currently on leave and serving as head of Business Intelligence Lab and Talent Intelligence Center at Baidu Inc. His research interests include data mining, mobile computing, and their applications in business. He has authored over 200 research articles, and co-edited or coauthored 4 books including the widely used Encyclopedia of GIS, which has been recognized as the Top 10 Most Popular Computer Science Book authored by Chinese scholars at Springer. Dr. Xiong has served as chair/co-chair for many international conferences in data mining, including a Program Co-Chair (2013) and a General Co-Chair (2015) for the IEEE International Conference on Data Mining (ICDM), and a Program Co-Chair of the Research Track (2018) and the Industry Track (2012) for the ACM SIGKDD International Conference on Knowledge Discovery and Data Mining (KDD). Dr. Xiong’s research has generated substantive impact beyond academia. He is an ACM distinguished scientist and has been honored by the 2018 Ram Charan Management Practice Award as the Grand Prix winner from the Harvard Business Review, the 2017 IEEE ICDM Outstanding Service Award, and the ICDM-2011 Best Research Paper Award.

Mobile Analytics: Prospects and Opportunities

Advances in sensor, wireless communication, and information infrastructure such as GPS, WiFi, and mobile phone technology have enabled us to collect and process massive amounts of mobile data from multiple sources but under operational time. These big data have become a major driving force of new waves of productivity growth, application innovation, and consumer surplus.  The big data are usually immense, fine-grained, diversified, dynamic, and sufficiently information-rich in nature, and thus demand a radical change in the philosophy of data analytics. In this talk, we discuss the technical and domain challenges of big data analytics in mobile environments. In particularly, it is especially important to investigate how the underlying computational models can be adapted for managing the uncertainties in relation to big data process in a huge nebulous environment.  The theme to be covered will include AI enabled map services (e.g. multi-modal travel recommendation), context-aware POI recommendations, POI knowledge graph, and urban cognitive computing.

青年报告人

按报告顺序排序

陶表帅

Recent Advancements in the Theory of Influence Maximization

Abstract

People often adopt improved behaviors, products, or ideas through the influence of friends.  One way to spread such positive elements through society is to identify those most influential agents, those that cause the maximum spread, and initiate the spread by seeding them. However, this strategy has a key difficulty: finding these influential seed nodes. This is difficult even if both the network structure and the way how the spread goes are known. In social networks, cascades model the phenomena that agents receive certain information, adopt certain products, or take up certain political opinions from their neighbors due to their influence. In the influence maximization problem, a central planner is given a graph and a limited budget k, and he needs to pick k seeds such that the expected total number of infected vertices in the graph at the end of the cascade is maximized. This problem plays a central role in viral marketing, outbreak detection, rumor controls, etc.

This talk will focus on computational complexity, approximability and algorithm design aspects of the influence maximization problem, with both submodular and nonsubmodular cascade models. 

Biography

Biaoshuai Tao is currently working toward the Ph.D. degree with the Computer Science and Engineering Division, University of Michigan, Ann Arbor, MI, USA, and he is expected to graduate in spring, 2020. His research interests mainly include the interdisciplinary area between theoretical computer science and economics, including social network analyses, resource allocation problems and algorithmic game theory. Before joining the University of Michigan, Biaoshuai was employed as a project officer at Nanyang Technological University in Singapore from 2012 to 2015, and he received the B.S. degree in mathematical science with a minor in computing from Nanyang Technological University in 2012.

朱瑞禹

Salable Actively-Secure Multiparty Computations

Abstract

Secure computation has long been speculated to be a key technology for safely utilizing sensitive data owned by two or more distrustful parties. Towards this goal, a number of theoretical and implementational breakthroughs have signicantly advanced the practicality of secure computation. However, even the state-of-the-art offer exciting performance numbers in the experiments, they all

suffer from either a high roundtrip latency or a prohibitive memory demand once scaled up. In this talk, we are going to see how to address these concerns in both 2-party setting and multi-party setting with lightweight augments and almost zero performance overhead, and scales secure computation to an unprecedented scale.

Biography

Ruiyu Zhu is an active researcher in applied cryptography. He has been publishing continuously as the author to top-tier conferences. He has also served on workshop committee and as top journal reviewers. His research focuses on practical multi-party computations.

庄   勇

Large-scale Machine Learning: a Case Study of Matrix Factorization and Field-aware Factorization Machines

Abstract

The rise of machine learning in the age of data has been evident for years now. With the growth of data, we should keep proposing new algorithms to effectively learn from data. In addition, given existing algorithms, with the growth of computational power and resource, it is also important to focus on how to train existing models more efficiently. In this presentation, I will first introduce a fast parallel SGD method, FPSGD, for matrix factorization in shared memory systems to demonstrate how to learn from data efficiently. Then, I will use CTR Prediction as a use case to demonstrate how to use field aware factorization machines (FFMs) to learn from data effectively.

Biography

Yong Zhuang is a Ph.D. of Electrical and Computer Engineering (ECE) at Carnegie Mellon University (CMU). Before joining the Ph.D. program of CMU, he received his master degree in Computer Science at National Taiwan University under supervision of Prof. Chih-Jen Lin, and his bachelor degree in Computer Science at Zhejiang University. Yong Zhuang's research interests lie in applied machine learning, and networks science. Specific research topics include the recommender system, computational advertising,  fast (parallel) training for large-scale ML models, and information networks. Yong Zhuang won the best paper award in ACM RecSys'13, the first prize of both tracks in KDD Cup 2013, the first prize of Criteo Display Advertising Competition, and the first prize of Avazu Click-Through Rate Prediction. His highest ranking on Kaggle is 30.

蒋才桂

Geometry, Architecture, and Fabrication

Abstract

Geometry plays an essential role in modeling and fabrication of complex objects from small-scale industry products to large-scale architecture. In the past decade, the arising new research area of architectural geometry at the interface of Mathematics, Computer Science, Structural Engineering and Architecture, sought to obtain a balance between purely freeform design and fabrication efficiency. In my presentation, I will briefly introduce several of my recent projects on architectural geometry, including shading and lighting system, freeform honeycomb structures, freeform space truss design, and curved-pleated structures. These projects provide useful insights for modern architecture design and illustrate the beauty of geometry behind.

Biography

Caigui Jiang is currently a research scientist at the Visual Computing Center (VCC) of King Abdullah University of Science and Technology (KAUST). Before that, he worked as a postdoctoral researcher at the Max Planck Institute for Informatics and ICSI of UC Berkeley from Jun. 2016 to Jan. 2019. He obtained his Ph.D. from KAUST under the supervision of Prof. Dr. Helmut Pottmann and Prof. Dr. Peter Wonka, and B.S. and M.S. degrees from Xi'an Jiaotong University (XJTU) in 2008 and 2011 respectively. His interests lie in the interface of applied geometry, computer science, structural engineering, and architecture. He has published in SIGGRAPH/ Asia, ACM TOG, Eurographics, SGP, and Advances in Architectural Geometry, where he received the Best Paper Award 2016.

秦   宸

Deep Learning for Magnetic Resonance Image Reconstruction and Analysis

Abstract

Recent advances in deep learning have shown great potentials in improving the entire medical imaging pipeline, from image acquisition and reconstruction to disease diagnosis. In this talk, I will introduce several deep learning techniques for medical imaging, and will mainly focus on Magnetic Resonance (MR) image reconstruction and analysis. Firstly, I will introduce my recent study on dynamic MR image reconstruction from highly undersampled k-space data. A CRNN (convolutional recurrent neural network) model will be presented which models the traditional iterative optimisation process in a learning setting and is able to exploit the spatio-temporal redundancies effectively and efficiently. As a complement, a k-t NEXT (k-t Network with X-f Transform) method will be introduced in which image signals are recovered by alternating the reconstruction process between x-f space and image space in an iterative fashion. Secondly, I will present some of my research on MR image analysis including both image segmentation and image registration. For image segmentation, an uResNet (u-shaped residual network) approach will be introduced to address the differential segmentation of white matter hyperintensities and stroke lesions on brain MRI. For image registration, I will go through my latest work, an unsupervised multi-modal deformable image registration (UMDIR) method, where we proposed to address the multi-modal registration problem by reducing it to a mono-modal one via disentangled representations. Finally, a joint framework which simultaneously predict cardiac motion estimation and segmentation will be presented, where it shows that the segmentation and registration tasks are beneficial from each other.

Biography

Chen Qin is now a final-stage PhD student at Biomedical Image Analysis Group in the Department of Computing, Imperial College London under the supervision of Prof. Daniel Rueckert. She will be starting as a post-doctoral researcher at Imperial College London from October, 2019. Before that, she received her Master of Science degree in Control Science and Engineering at Tsinghua University. Her PhD thesis mainly focuses on machine learning for magnetic resonance image (MRI) reconstruction and analysis, and she has been working on investigating machine learning especially deep learning methods for medical image analysis, including dynamic MRI reconstruction, image registration and image segmentation. She has also worked as a research intern at Siemense Healthineers at Princeton and Huawei London Research Center. During her doctoral research, she has published around 20 papers on top-tier journals and conferences such as IEEE Transactions on Medical Imaging (TMI), International Conference on Medical Image Computing and Computer-Assisted Intervention (MICCAI), and International Conference on Information Processing in Medical Imaging (IPMI). Among them, 10 papers are first authored by Chen, with 4 journal and 6 conference articles. She has also been invited to speak at international conferences such as International Society for Magnetic Resonance in Medicine (ISMRM), International Biomedical and Astronomical Signal Processing (BASP) Frontiers workshop etc. Besides, Chen Qin also serves as a reviewer for several journal and conference publications including MedIA, MRM, JBHI and MICCAI.

王大阔

A New Paradigm Of Human-Computer Interaction

Abstract

We are witnessing an emerging new paradigm of how people interact with computer systems. A few examples include computers we ate into our stomachs, artificial limbs controlled by electrical muscle stimulation (EMS), e-tattoos on our skins, various Augmented Reality/Virtual Reality (AR/VR) systems, auto-piloted cars, and various smart assistants in our phones, at work, and at home. Contrary to the known paradigms, where we know how people use Personal Computers (PCs) with Graphical User Interfaces (GUI), web-based systems, mobile applications (Ubiquitous Computing), social network systems (Social Computing), we have little knowledge about how people interact with these new technologies. We know so little that we do not even have a consensus of the name for this new paradigm, thus the design of such systems is quite opportunistic. In this talk, I will firstly provide an account and a definition for this new paradigm. Then, I propose a framework, which can be used as a guideline for understanding people's interactions with new and smart systems, and as design principles for designing such systems.

周鹏展

Data-driven Online Optimization of Parking Placement and Truthful Incentivizing for Electric Bike Sharing

Abstract

The rise of dockless electric bike sharing offers a new modality of green transportation while also brings new challenges to urban management and maintenance. Due to the safety risks of batteries, customers are regulated to park at designated locations, which potentially causes dissatisfaction and customer loss. Meanwhile, service providers should charge those scattering low-energy batteries in time. To address these issues, we propose E-sharing, a two-tier optimization framework that leverages data-driven online algorithms to plan parking locations and maintenance. We also explore truthful incentivizing mechanism to seek user cooperation to re-balance the e-bikes by applying reinforcement learning for a general system with k levels of difficulty.

Biography

Pengzhan Zhou received the B.S. degree in both Applied Physics and Applied Mathematics from Shanghai Jiaotong University, Shanghai, China. He is currently working towards the PhD degree at the Department of Electrical and Computer Engineering, Stony Brook University, New York. His research interests include artificial intelligence, computational economy, combinatorial optimization, incentivizing mechanisms and algorithms.

▽▼ 听众报名现在开放 ▼▽

听众报名截止：2019年10月21日17:00，扫描下方小程序码报名

听众报名确认：2019年10月22日

论坛时间：2019年10月23日

↓↓↓ 扫码报名 ↓↓↓

联系人：杨老师  [e-mail]

想直接申请中心职位？招聘广告直通车——

（广告长期有效）

往 届 论 坛 回 顾

—   版权声明  —

本微信公众号所有内容，由北京大学前沿计算研究中心微信自身创作、收集的文字、图片和音视频资料，版权属北京大学前沿计算研究中心微信所有；从公开渠道收集、整理及授权转载的文字、图片和音视频资料，版权属原作者。本公众号内容原作者如不愿意在本号刊登内容，请及时通知本号，予以删除。

点击“阅读原文”转网页版论坛通知