Keynote Speaker 1:
Prof. Assaf Schuster
ACM Fellow&IEEE Fellow, Israel Institute of Technology, Israel
Speech Title: Distributed Training of Deep Neural Networks
Abstract: Modern deep neural networks are comprised of millions and billions of parameters, which require massive amounts of data and time to train. Steady growth of these networks along the years has made it impractical to train them from scratch on a single GPU. Distributing the computations over several GPUs can drastically reduce the training time, however, stochastic gradient descent (SGD), which is typically used to train these networks, is an inherently sequential algorithm. As a result, training deep neural networks on multiple workers is difficult, especially when using non-dedicated cloud resources trying to maintain high efficiency, scalability and final accuracy. I this talk we will survey some of the new ideas in this scope and discuss their potential.
Biography: Prof. Assaf Schuster of the Computer Science Department is the head of the new AI center at the Technion. He is a Fellow of the ACM and the IEEE, with more than 200 published papers in highly selected venues. His interests and publications are in the wide scope of distributed and scalable data mining, big and streaming data technologies including management, analytics & prediction, cyber security and system/IoT vulnerabilities, privacy preserving, cloud resource management and more. He consulted leading hi-tech companies and participated in the bumpy journey of several startups, two of which he co-founded.
Keynote Speaker 2:
Prof. Xu Lei,
Fellow of IEEE, IAPR Fellow, The Chinese University of Hong Kong, China
Speech Title: Reasoning and Casual Computation from A Bidirectional Intelligence Perspective
Abstract: After a brief overview on bidirectional learning studies from the later eighties and the early nineties (e.g., autoencoder, Lmser, etc) to recent years (VAE, GAN, U-net, etc), we proceed to bidirectional intelligence, driven by long term dynamics for parameter learning and short term dynamics for image thinking and rational reasoning, including topics of (a) Why deep learning and why Lmser from a viewpoint of holistic inference vs DAG reasoning, (b) learning dependence versus discovering causal relation; (c) three levels and bidirectional framework for reasoning, (d) causal potential theory and its enhanced rho-diagram equations for DAG causal analyses.
Biography: Lei Xu, Emeritus Professor, Chinese University of Hong Kong; Zhiyuan Chair Professor, Shanghai Jiao Tong University (SJTU); Chief Scientist of SJTU AI Research Institute， and of SJTU-Sensetime Research Institute; Director of Neural Computation Research Centre in Brain and Intelligence Science-Technology Institute, ZhangJiang National Lab; Received several national and international academic awards, including 1993 National Nature Science Award, 1995 Leadership Award from International Neural Networks Society (INNS) and 2006 APNNA Outstanding Achievement Award. Elected to Fellow of IEEE in 2001; Fellow of intl. Association for Pattern Recognition in 2002 and of European Academy of Sciences (EURASC) in 2003. Published about 100 Journal papers, given over dozens keynote /invited lectures at various international conferences. Served as EIC and associate editors of several academic journals, e.g., including Neural Networks (1995-2016), IEEE Tr. Neural Networks (1994-98). Taken various roles in academic societies, e.g., INNS Governing Board (2001-03), the INNS award committee (2002-03), and the Fellow committee of IEEE Computational Intelligence society (2006-07), and the EURASC scientific committee (2014-17).
Plenary Speaker 1:
Prof. Lars Lundberg
Biography: Lars Lundberg is since 20 years a full professor of computer systems engineering at Blekinge Institute of Technology (BTH) in Sweden. Lars took has a master in computer science from Linköping university (Sweden), and a Ph.D. from Lund University (Sweden). Professor Lundberg has had a number of tasks at BTH, including being the dean of the technical faculty for six years, and the head of the department of computer science and engineering. Currently he is the dean of the faculty of computing at Blekinge Institute of Technology. Professor Lundberg has a long track record of working with industry in Sweden, USA, India and China. Lars has advised 14 PhD students to their doctoral degree and published more than 200 papers in international journals and conferences. His research interests include machine learning, real-time systems, high-performance processing, software engineering and cloud computing.
More information will be updated soon...