PROGRAMME

Title: Applications of AI in Healthcare
The increasing availability of new diverse sources of data as well as AI techniques to exploit such diverse data are coming together to transform the way that information is used to support medical and public health practice. In this talk I illustrate this trend in medicine and public health through our research in the areas of medical training and public health decision support.
The teaching of surgical skills, which has traditionally followed an apprentice style approach, is facing a number of challenges that are driving medical schools to seek alternatives. Medical schools are increasingly introducing simulators into their curricula as a way to provide students with increased practice. The deployment of such systems provides a great opportunity to gather data on student performance and to use this data to personalize instruction. I describe our work on intelligent environments for surgical training that seeks to explore the potential of the integration of techniques from Virtual Reality and Intelligent Tutoring Systems to increase the effectiveness of surgical training by providing personalized instruction, assessment, and formative feedback in a form and on a scale not possible in the physical world. I describe techniques to train physical skills and to train decision making skills.
Effective control of infectious disease requires rapid intervention in order to prevent spread from local areas of transmission. Since many diseases begin in relatively remote areas where public health resources are scarce, targeted intervention and resource allocation become crucial. Disease prediction models can help support such decisions. I describe our work on spatiotemporal Bayesian networks for malaria prediction, which are able to model the complex relationships between environment and disease necessary for accurate prediction. I present techniques to automatically generate models tailored to the specifics of the geographic region being modeled. I then discuss the relationship between spatial resolution and prediction accuracy, which is well known but poorly studied. I present complexity-based spatial hierarchical clustering, a new technique that effectively exploits the tradeoff and provides guidance in selection of appropriate spatial resolution.
Speaker Biography Professor Haddawy received MSc and PhD degrees in Computer Science from the University of Illinois-Urbana in 1986 and 1991, respectively. He has been a Fulbright Fellow, Hanse-Wissenschaftskolleg Fellow, Avery Brundage Scholar, and Shell Oil Company Fellow. His research falls broadly in the areas of Artificial Intelligence, Medical Informatics, and Scientometrics and he has published over 120 refereed papers with his work widely cited. His research in Artificial Intelligence has concentrated on the use of decision-theoretic principles to build intelligent systems and he has conducted seminal work in the areas of decision-theoretic planning and probability logic. His current work focuses on intelligent medical training systems and modeling of vector-borne disease. In the area of Scientometrics Prof. Haddawy has focused on development of novel analytical techniques motivated by and applied to practical policy issues. He currently holds a professorship in the Faculty of ICT at Mahidol University in Thailand.

Title: Deep Learning for Computer Vision
Abstract: Recently deep learning has become one of the most powerful and popular machine learning techniques due to its record-breaking performances in a variety of recognition tasks including speech recognition and image classification. Deep learning also changed the paradigm of pattern recognition in that it allows us to automatically discover hierarchical features from data instead of relying on hand-crafted features. In this tutorial, I will provide an overview of deep learning discussing what have been the main difficulties of training deep neural networks and how these difficulties have been overcome by recent breakthroughs. I will also introduce several deep learning techniques such as restricted Boltzmann machine (RBM), deep belief network (DBN), deep neural network (DNN), and convolutional neural network (CNN) and talk about how they are applied to computer vision problems such as a large scale image classification.
Speaker Biography Dr. Junmo Kim received the B.S. degree from Seoul National University, Seoul, Korea, in 1998, and the M.S. and Ph.D. degrees from the Massachusetts Institute of Technology (MIT), Cambridge, in 2000 and 2005, respectively. From 2005 to 2009, he was with the Samsung Advanced Institute of Technology (SAIT), Korea, as a Research Staff Member. He joined the faculty of KAIST in 2009, where he is currently an Associate Professor of electrical engineering. His research interests are in image processing, computer vision, statistical signal processing, machine learning, and information theory.

Title: Extreme Learning Machines (ELM) – Filling the Gap between Machine Learning and Biological Learning
Abstract: One of the most curious in the world is how brains produce intelligence. Brains have been considered one of the most complicated things in the universe. Machine learning and biological learning are often considered separate topics in the years. The objectives of this talk are three-folds: 1) It will analyse the differences and relationships between artificial intelligence and machine learning, and advocates that artificial intelligence and machine learning tend to become different, they have different focus and techniques; 2) There exists some convergence between machine learning and biological learning; 3) Although there exist many different types of techniques for machine learning and also many different types of learning mechanism in brains, Extreme Learning Machines (ELM) as a common learning mechanism may fill the gap between machine learning and biological learning, in fact, ELM theories have been validated by more and more direct biological evidences recently. ELM theories actually show that brains may be globally ordered but may be locally random. ELM theories further prove that such a learning system happens to have regression, classification, sparse coding, clustering, compression and feature learning capabilities, which are fundamental to cognition and reasoning. This talk also shows how ELM unifies SVM, PCA, NMF and a few other learning algorithms which indeed provide suboptimal solutions compared to ELM.
Speaker BiographyGuang-Bin Huang is a Full Professor in the School of Electrical and Electronic Engineering, Nanyang Technological University, Singapore. He is a member of Elsevier's Research Data Management Advisory Board. He is one of three Expert Directors for Expert Committee of China Big Data Industry Ecological Alliance organized by China Ministry of Industry and Information Technology, and a member of International Robotic Expert Committee for China. He was a Nominee of 2016 Singapore President Science Award, was awarded by Thomson Reuters “Highly Cited Researcher” (in two fields: Engineering and Computer Science), and listed in Thomson Reuters’s “The World's Most Influential Scientific Minds.” He received the best paper award from IEEE Transactions on Neural Networks and Learning Systems (2013). His two works on Extreme Learning Machines (ELM) have been listed by Google Scholar in 2017 as Top 2 and Top 7, respectively in its “Classic Papers: Articles That Have Stood The Test of Time” - Top 10 in Artificial Intelligence.

Title: Rule Based Fuzzy Systems and Applications
1. Basic concepts, operations and relations;
2. Classic fuzzy rule based system;
3. Decision making and control in fuzzy signature set rule bases; and
4. Example of applications.

Speaker Biography Laszlo T. Koczy received the M.Sc., M.Phil. and Ph.D. degrees from the Technical University of Budapest (BME) in 1975, 1976 and 1977, respectively; and the D.Sc. degree from the Hungarian Academy of Science in 1998. He spent his career at BME until 2001, and from 2002 at Szechenyi Istvan University (Gyor, SZE). He has been from 2002 to 2011 Dean of Engineering, and from 2013 to current President of the University Research Council and of the University Ph.D. Council. From 2012 he has been a member of the Hungarian Accreditation Committee (for higher education), appointed by the Prime Minister, and elected Chair of the Engineering and Computer Science sub-committee, member of the Professors and Ph.D. sub-committee, and has been a member of the National Doctoral Council since 2012.