SS1 - Intelligent and Expert Systems with Industrial Applications
Organizer: Dr. Justin Pang Chee Khiang, National University of Singapore, Singapore
In general, intelligent information processing attempts to utilize advanced knowledge processing methodologies for extracting and elucidating useful information from various sources, which include data, text, as well as audio and visual signals. In this context, fuzzy logic provides a useful means to design and develop effective models and tools for intelligent information processing. Introduced by Prof. L.A Zadeh in 1965, fuzzy logic constitutes a form of multi-valued logic, and is capable of dealing with the concept of partial truth, imprecision, and vagueness. It has become a popular and important methodology which is able to exploit the approximate reasoning capability of the human mind and the underlying information from human linguistic variables in undertaking various challenging problems.
Supply chain and logistics systems in the current age are complex networks as the result of globalization, outsourcing and lean initiatives. The increasing interests in complex systems are driven predominantly by new trends, challenges and demands in practical systems such as supply chain and logistics systems where industries have the interest on how to design, manage, build and control systems as they increase in scale and complexity. They hope to build systems that are scalable, robust, and adaptive by using properties such as self-organization, self-adaptation, and manage the disruptions happening in their systems.
Today, Swarm Intelligence has become an increasingly important driving role for business strategies and real modern world applications. The essential idea of Swarm Intelligence is the emergent collective intelligence of groups of simple agents. There are several popular algorithms based on these concepts, including Particle Swarm Optimization (PSO), Ant Colony Optimization (ACO), Artificial Bee Colony (ABC) algorithms and Simplified Swarm Optimization (SSO), etc. Despite a significant amount of research on Swarm Intelligence, there remain many open issues and intriguing challenges in the field. The aims of this special session are to demonstrate the current state-of-the-art concepts, theory and practice of Swarm Intelligence, to reflect on the latest advances in swarm intelligent design and applications for real world problems, and to explore the future directions in Swarm Intelligence.
This special session attempts to forge a discussion on the recent developments and the future directions of Agent-Based Computational Economics, in the light of advances in Experimental Economics. In particular, this calls for a greater attention to the design of artificial agents for which the findings from Experimental Economics might prove to be quite insightful. Agent Based Modeling (ABM) has been steadily gaining more importance in the field of Economics and Finance. It is being viewed as an alternative approach to the conventional equation-based economic modeling, and as a useful methodology to investigate complex systems around us. In its early stages, developments in ABM paralleled those in Computational Intelligence, such as Zero-Intelligence Agents, Cellular Automata, Artificial Neural Networks, Fuzzy Logic, Swarm Intelligence, Genetic Algorithm, and Genetic Programming. It marked a truly interdisciplinary approach to social sciences, drawing insights from a variety of fields such as mathematics, biology, neuroscience, and computer science among others. Today, the development of ABM has reached a stage where the evidence concerning human decision making, largely arising from lab and filed experiments can no longer be neglected if it were to provide a formidable alternative. Among the factors affecting human decision making, cognitive capacity, personality, culture, emotion and social preference are well recognized and documented. However, integrating these attributes within the existing models of artificial agents is far from being trivial.
There are many real network such as Internet, social networks, citation networks,infection network, diffusion network, and bank networks. And many scientists havemodeled and developed methods to help us understand the behavior of systems. In this field, we seek to explain models of network growth and dynamical processes taking place on networks. It is said that there is much to be done by M. E. J. Newman in this field. First, while we are beginning to understand some of the patterns and statistical regularities in the structure of real world networks, our techniques for analyzing networks are at present no more than a grab-bag of miscellaneous and largely unrelated tools. Second, there is much to be done in developing more sophisticated models of networks, both to help us understand network topology and to act as a substrate for the study of processes taking place on networks.
The Internet is the largest network. The Internet architecture is fundamentally a “client-server" architecture, with limited service capability and static routing/addressing. The intriguing interplay of local and topological dynamics makes robust self-organization possible in these networks. Recently important breakthroughs in the understanding of adaptable networks have been made and it is becoming more apparent that they could hold the key to many phenomena observed in a wide variety of applications. The goal of this special session is to promote research and development in Intelligent network which therefore presently receive rapidly increasing attention from researchers from very different fields including Genetic Algorithm, Artificial Intelligent, Practical Swarm Optimisation, Evolutionary Multi-objective Optimisation Algorithms, Evolutionary Algorithm, etc. The next wave of research in the field of Internet Architecture should solve remaining problems and bring the most promising options closer to deployment.
In today’s modern life, societies have to embrace digital technologies in their homes and use them to enhance the quality of life. The emerging of aging societies around the world prompts more urgent need for having more affordable and easy to set up “smart home systems”, as well as assisted technologies used inside a home with high performance and efficient solutions and platforms. Adoption of smartphones, smartTVs, and other smart devices, adds ubiquity and mobility to existing home-based self-care and health monitoring systems and expands the horizon of endless possibility in research and development.
Memetic Computing (MC) represents a broad generic framework using the notion of meme(s) as units of information encoded in computational representations for the purpose of problem-solving. MC usually emerges as population-based meta-heuristic algorithms or hybrid global-local search, inspired by Darwinian principles of natural selection and Dawkins’ notion of a meme defined as a unit of cultural evolution that is capable of local/individual refinements. Taking advantage of both biological selection and cultural selection, a plethora of potentially rich MC methodologies, frameworks and operational meme-inspired algorithms have been developed with considerable success in various real-world applications. Yet, there remain many open questions emerging as intriguing challenges for the field. The goal of this session is to serve as a forum for reseachers in this field to exchange the latest advances in theories, technologies, and practice of MC.
Computational intelligent Techniques are attractive global optimization methods inspired by the various phenomena arising in nature and man-made problems. They include Bat algorithm, Teaching and Learning based algorithm, Artificial Neural Networks, Fire fly, Simulated annealing, Tabu search, Variable neighborhood search, Support vector machine, Genetic Algorithms, Memetic Algorithms, Differential Evolutions, Particle Swarm Optimization (PSO), Glowworm Swarm Optimization, Bee Algorithms, Bacterial Foraging, Ant Colony Algorithms, Chaotic algorithm etc. These are relatively a newer addition to the class of numerical optimization algorithms. These methods have been successfully applied to a wide range of real-world application problems. Natural disasters and made chaotic problems partially can be solve by classical optimization techniques. Modern optimization algorithms are capable to handle and tackle these problems with higher level of degree of satisfaction.