Instructor: Siba K. Udgata (University of Hyderabad, India)
Time and venue: 9:30-12:00, December 8, Meeting Room, Faculty of Informatics building
Description
Swarm intelligence is an emerging field of biologically-inspired artificial intelligence based on the behavioral models of social insects such as ants, bees, wasps and termites (Bonabeau, 1999). Swarm intelligence provides a basic scientific theory about how complex and sophisticated behaviors can emerge from social creature groups.
To tackle complex real world problems, scientists have been looking into various new paradigms, natural processes and creatures both as model and metaphor. Optimization is at the heart of many natural processes including Darwinian evolution, social group behavior and foraging strategies. Over the last few decades, there has been remarkable growth in the field of nature-inspired search and optimization algorithms.
Currently swarm intelligence techniques are applied to a variety of problems, ranging from scientific research to industry and commerce. These domains extend to Aerospace, Communication, Computer Networks, Systems and Control, Robotics, Power Systems, Financial services and Engineering, Management Information Systems, Business Intelligence, WWW and Search, Image Processing, Computational Science, Chemistry, Structural and Mechanical Designs, Bioinformatics, Environmental Modeling and Software Engineering and many others.
There are various swarm intelligence techniques proposed in the literature, each one having its own advantages and limitations. Particle Swarm, Ant Colony, Bacterial Foraging, Artificial Bees, Fireflies, Invasive Weed Optimization, Harmony Search are a few popular and potential swarm intelligence techniques proposed in the past and have been applied to various problem domains. Many variants and hybridization of these algorithms also have been proposed for better performance.
In this tutorial, it is proposed to give a brief overview of these swarm intelligence algorithms. The talk will also have an in-depth comparative study of these algorithms in different domains. This will focus on the guiding principle of different techniques ad their suitability for different problems. In particular, it will focus on some IT related applications like clustering in data mining, routing in networks, node deployment in wireless sensor networks, finding shortest path for packet forwarding, optimum resource allocation and planning, software failure prediction in software engineering among many others.
It is also proposed to have some hands-on (particularly for students) experience for writing the various swarm intelligence algorithm codes in a MATLAB environment and have a feel of the working principles. The session can have a few demonstrations about (i) developing the code from the algorithm, (ii) how to tune different parameters for optimal results, (iii) application of swarm intelligence algorithms in small problems of different domains of interest to participants.
About the instructor
He has published extensively in different refereed international journals and conferences in India as well as abroad. He has visited many countries like Hong Kong, China, Japan, United States of America, Singapore and Nepal. He was a United Nations Fellow and worked in the United Nations University/ International Institute for Software Technology (UNU/IIST), Macau in the year 2001.
He was a visiting Professor at Nepal Engineering College, Nepal and Mahasarakham University, Thailand. He was also a visiting fellow at Ball State University, Muncie, Indiana, USA.
Swarm Intelligence -
A Tool for Optimization in various Information Technology Applications
The 4th Mahasarakham International Workshop on Artificial Intelligence
December 9-10, 2010, Mahasarakham University, Mahasarakham, Thailand
Sponsors