Keynotes
Unfortunately the keynotes by Suzanne Gildert and Geordie Rose have been cancelled due to last-minute scheduling conflicts.
____________________________________________________________________________
Wei Cui, Ph.D., Co-founder & Chief Scientist, Squirrel AI Learning (Shanghai)
Biography:
Dr. Cui had worked in top European adaptive education enterprise for many years, and was responsible for researching on adaptive learning algorithms and systems. His achievements in adaptive learning system had been popularized among European secondary and higher education market.
Dr. Cui has published 16 international academic papers including one in Soft Computing, which is the core journal of artificial intelligence, one in Quantitative Finance, which is the top periodical all over the world, and one in IEEE-CiFEr as the best conference paper. Dr. Cui has delivered 19 speeches at Internationally-renowned academic conferences. He also founded the first Chinese online shopping platform in Ireland, covering 80% of Chinese in Ireland.
Dr. Cui has received a bachelor’s degree in software engineering from Wuhan University, a master degree in communication engineering, and the doctor’s degree in Artificial Intelligence of the National University of Ireland.
____________________________________________________________________________
Hugo Latapie, CTO NDS Americas, now a part of Cisco
Title: Beyond ML/DL: Learning By Reasoning For Smarter Cities And Safer Schools
____________________________________________________________________________
Zhongzhi Shi, Institute of Computing Technology, Chinese Academy of Sciences
Biography:
Zhongzhi Shi, Professor at the Institute of Computing Technology, Chinese Academy of Sciences. Fellow of CCF and CAAI. IEEE senior members, AAAI, ACM members. His research interests mainly contain intelligence science, artificial intelligence, multi-agent systems, machine learning. He has been responsible for 973, 863, key projects of NSFC. He has been awarded with various honors, such as National Science and Technology Progress Award (2012), Beijing Municipal Science and Technology Award (2006), the Achievement Award of Wu Wenjun artificial intelligence science and technology by CAAI (2013), the Achievement Award of Multi-Agent Systems by China Multi-Agent Systems Technical Group of AIPR, CCF (2016). He has published 16 books, including “Mind Computation”, “Intelligence Science”, “Advanced Artificial Intelligence”, “Principles of Machine Learning”. Published more than 500 academic papers. He served as chair of the machine learning and data mining group, IFIP TC12. He served as Secretary-General of China Computer Federation, vice chair of China Association of Artificial Intelligence.
Title: Progress In Research On Intelligence Science
Abstract:
Intelligence Science is an interdisciplinary subject which dedicates to joint research on basic theory and technology of intelligence by brain science, cognitive science, artificial intelligence and others. Brain science explores the essence of brain, research on the principle and model of natural intelligence in molecular, cell and behavior level. Cognitive science studies human mental activity, such as perception, learning, memory, thinking, consciousness etc. In order to implement machine intelligence, artificial intelligence attempts simulation, extension and expansion of human intelligence using artificial methodology and technology. At present, intelligence science is an active research area which aims to lead artificial general intelligence and the new generation of artificial intelligence. Brain-inspired research will mainly applied to the research and development of artificial general intelligence. In this talk, I will discuss mind models, cognitive machine learning and the cognitive model of brain machine integration.
____________________________________________________________________________
Harri Valpola, CEO Curious AI
Keynote 1: Thursday the 8th of August at 15:30 (for technical audience)
Title:
Real-world model-based reinforcement learning with deep neural networks
Abstract:
Reinforcement learning (RL) research has mainly focused on model-free techniques which rely on trial and error to learn a policy network. While this can yield impressive results with enough training, model-free RL techniques are nowhere as efficient learners as humans. As a result, these techniques have shined in simulation environments but are often impractical in the real world where trials and errors cost time and money. Model-based RL is known to be much more sample efficient, approaching human speeds of learning. It relies on learning an explicit causal model of the environment and then searching through alternative action plans to find the best actions. Unfortunately there have been several problems which have limited its use. In particular, it has been difficult to use MBRL with expressive models such as deep learning. I will discuss the root cause of the problem and demonstrate a solution which has allowed Curious AI to apply general MBRL techniques to real-world problems such as industrial process control and control of autonomous mining machines.
Keynote 2: Friday the 9th of August at 11:00 (for general audience)
Title:
Learning and imagination: crucial ingredients for AGI
Abstract:
While AI is now powering many real-world applications, it is still not the kind of general and autonomous AI which people have envisioned in science fiction. Instead, AI heavily relies on humans in defining goals, datasets, abstractions, etc. I will argue that this is because we haven’t managed to properly combine learning and imagination (internal simulations of the world). My talk on the previous day explains how model-based reinforcement learning can be implemented by expressive models such as deep neural networks. This exemplifies how learning and imagination can serve control and decision-making. In this talk I will discuss how these mechanisms underlie general intelligence, communication and learning in humans. I will also go through real-world examples where learning and imagination serve analysis tasks in AI.
____________________________________________________________________________
Wei Xu, Chief Scientist of General AI at Horizon Robotics
Title: Developing Machine Intelligence in an Interactive and Embodied Setting
Abstract: One of the key ingredients of intelligence is learning, that is, the ability to improve oneself through experience. A truly intelligent machine needs to be able to learn quickly from two types of experience: the experience of interacting with the physical environment, from which common sense knowledge of the physical world can be acquired, and the experience of interacting with human, from which human language and vast amount of human knowledge can be acquired. In order to develop agents that learn from these kind of experience, we need two types of infrastructure: rich simulation environments that agents learn from, and flexible framework which make it easier to design complex algorithms for such agents. I will introduce our on-going projects on these two fronts: SocialRobot simulation environment and ALF agent learning framework. I will also talk about several of our recent works on language learning in an interactive and embodied setting.
____________________________________________________________________________
Yi Zeng, Professor, Deputy Director, Research Center for Brain-inspired Intelligence, Institute of Automation, Chinese Academy of Sciences
Title: Creating Brain-inspired Conscious Living Becomings for Beneficial Human-AI Society
Abstract:
For the long term quest on the scientific nature of Artificial Intelligence, we need to explore the mechanisms of the biological brain and mind and use them as sources of inspirations to create computational models for cognitive intelligent machines and future computational conscious living becomings. In this talk, I will discusses Brain-inspired Artificial General Intelligence from scientific and philosophical point of views. I will introduce concrete models for multi-scale Brain-inspired learning, development, evolution and recent progress of brain-inspired self-consciousness models from my lab. We argue that future Brain-inspired Artificial Conscious Living Becomings are inspired by the mechanisms of the human brain and with embodiment to explore and interact with the nature and human society. Hence, it can be considered as a version and mirror to the current and future humanity. Through creating this kind of intelligent and conscious living becomings, we are not only in the process of the scientific quest for intelligence, but also in the process to understand the nature and shape the future of humanity. We need to realize that the future relies on how human and future brain-inspired artificial conscious living becomings shape the world together for a Beneficial Human-AI society.