ICT and Fusion
ICT refers to all types of technologies used to handle telecommunications, broadcast media, intelligent building management systems, audiovisual processing and transmission systems, and network-based control and monitoring functions. The potential and capabilities of modern ICT systems are still growing exponentially fueled by the progress in electronics, microsystems, networking, the ability to master increasingly complex cyber-physical systems and robots, and progress in data processing and human machine interfaces. This subject group include high performance computer (HPC), industrial mathematics for scientific computing which needs inter-discipline, image handling, big data, machine learning, and artificial intelligence (AI). At EKC 2017, three oral sessions and one poster session are planned for ICT and Fusion. 1) Industrial Mathematics for Scientific Computing: Imaging and Machine Learning 2) Industrial Mathematics for Scientific Computing: Computational Fluid Dynamics (CFD) 3) IOT and Big data 4) Poster presentation
Date / Time 2017-07-28 10:50   --   12:30
Room E6
Conveners / Chairs
  • DR. HAHN, Jooyoung (Software Development Engineer, AVL LIST GmbH) CONVENER
  • DR. HA, Taeyoung (NIMS) COCONVENER
Machine learning has been attracting attention as a method to solve many problems in various fields. The development of deep learning methods and its numerous applications, for instance, artificial intelligence are enough to attract the attention of many researchers. The research and development of machine learning algorithms, and findings of its applied fields are being expanded not only in the field of science but also around life, making it a convenient service for living. Machine learning algorithms can be used to a modelling method for highly complex fluid problems or detailed chemical kinetic mechanisms in the combustion process. It is also used in autonomous automobiles in combination with object recognition, speech recognition, sensor data analysis, etc., and is also being introduced into medical care for human life extension, health care for human welfare. Recently, machine learning has been suggested as a solution to solve industrial productivity and efficiency issues. We can find much more applications and usefulness of machine learning in many other areas. Imaging and image processing based on mathematics are widely used for solving problems in medicine and industry. Mathematics plays a key role in computational sciences as well as providing mathematical modelling and rigorous verification. However, although mathematics has proposed various potential methods through the modelling and verification, their direct application to medicine and industry has a room for improvement. Recently, machine learning techniques such as deep learning have been successfully applied to pattern recognition, segmentation, super-resolution, denoising and etc. Therefore, image processing based on deep learning is a promising methodology to improve diagnostic performances in the area of medicine. In this session, we will bring together researchers to present results related to deep learning for solving the current issues arising from various industrial and medical area.
  • PROF. PARENTE, Alessandro (Université Libre de bruxelles) [ 10:50 - 11:30 ]
    Title: Reduced-order modelling of turbulent reacting flows using principal component analysis
  • MR. HYON, Yunkyong (National Institute for Mathematical Sciences) [ 11:30 - 11:45 ]
    Title: Predicting materials properties using machine learning
  • DR. JANG, Jaeseong (National Institute for Mathematical Sciences) [ 11:45 - 12:00 ]
    Title: Neural network-based framework for automatic abdominal circumference measurement in fetal ultrasound
  • DR. PARK, Hyoung Suk (National Institute for Mathematical Sciences) [ 12:00 - 12:15 ]
    Title: Towards metal artifact reduction for X-ray computed tomography
  • DR. PARK, Won-kwang (Kookmin University) [ 12:15 - 12:30 ]
    Title: A study on direct sampling method for retrieving multiple targets
Date / Time 2017-07-28 13:30   --   15:10
Room E6
Conveners / Chairs
  • DR. KANG, Kab Seok (HLST Core Team, IPP) CONVENER
  • DR. HAHN, Jooyoung (Software Development Engineer, AVL LIST GmbH) COCONVENER
The computational fluid dynamics (CFD) has been one of crucial joint research topics between academia and industries. It is also one of topics to be extensively used to improve human life such as stability of commercial flights, low energy consumption of cars, high speed trains, etc. In mathematics, all kinds of analysis, topological, or geometrical tools are applied to understand highly complicated governing equations and it sometimes makes a driving force to develop new area of mathematics. Computational mathematics brings the numerical realization of partial differential equations in CFD. Such a realization in computer simulation is interpreted by engineering and physical methods and it usually makes a feedback of designing more realistic models to improve many industrial problems related to our real life. In this session, we would like to catch a small glimpse of the vast area in CFD from computational and engineering point of views. In multiphase flows, two specific examples, phase-change and stratified flows, are explained by using a state-of-the-art ultra-high resolution two-phase flow solver called TPLS backed up with fundamental theoretical analysis on instabilities, and advanced experimentation. Based on modern computer architectures, a rectangular parallelopiped domain for an elliptic partial differential equation is considered and a modification of the parallel multigrid algorithm is presented and it is implemented using MPI, by adding OpenMP parallelization at the node level. In combustion, numerical and theoretical analyses of the nonlinear dynamics of harmonically-forced turbulent premixed flames, using G-equation are presented. A key objective of this work is to develop a model that enables tracking of ensemble-averaged turbulent flame fronts. In numerical analysis and applications, shallow water equation on surfaces and G-equation are presented with recently developed numerical algorithms.
  • DR. VALLURI, Prashant (The University of Edinburgh) [ 13:30 - 14:10 ]
    Title: Beautiful multiphase flows at ultra-high resolution calculated at super-fast speeds: phase change applications and stratified flows
  • DR. KANG, Kab Seok (IPP) [ 14:10 - 14:25 ]
    Title: Multigrid method on KNL
  • DR. SHIN, Dong-hyuk (University of Edinburgh) [ 14:25 - 14:40 ]
    Title: Theoretical and numerical analysis of turbulent flame front dynamics
  • DR. LEE, Seunggyu (National Institute for Mathematical Sciences) [ 14:40 - 14:55 ]
    Title: Numerical simulation of the shallow water equation on a surfaces
  • DR. HAHN, Jooyoung (AVL LIST GmbH) [ 14:55 - 15:10 ]
    Title: Surface evolution in a polyhedron mesh
Date / Time 2017-07-27 16:30   --   18:10
Room E3
Conveners / Chairs
  • MR. KANG, Kab Seok (HLST Core Team, IPP) CONVENER
IoT and Big Data are hot issues in research and industry. These areas need to handle huge amount of data which require huge computing power and innovative algorithms which may come from AI. In this session, we share the current development and issues on IoT, Big Data, AI, and HPC in aspect of software and hardware.
  • DR. KANG, Kab Seok (IPP) [ 16:30 - 16:50 ]
    Title: Trends of the HPC and hardware for Big Data
  • DR. SUH, Hyun Kwon (Wageningen University & Research) [ 16:50 - 17:10 ]
    Title: Smart agriculture in the Netherlands and Northwest Europe
  • PROF. LEE, Youngjo (The Korean Academy of Sceince and Technology(KAST)) [ 17:10 - 17:30 ]
    Title: The Real Time Tracking and Alarming the Early Neurologic Deterioration using Blood Pressure Monitoring in Patient with Acute Ischemic Stroke
  • DR. CHOI, Jung Han (Fraunhofer Institute) [ 17:30 - 17:50 ]
    Title: Low-Power and High-Speed IC and Packages for optical communication
  • DR. SUNG, Ki Won (KTH Royal Institute of Technology) [ 17:50 - 18:10 ]
    Title: A cooperative system concept for broadcast and unicast in UHF band