Prof. Habiba Drias
University of Sciences and Technology Houari Boumediene, Algiers, Algeria
This course introduces the theoretical foundations of swarm intelligence as well as the practical features for developing concrete applications. Swarm intelligence aims at modeling the behaviors of particles such as ants, bees and birds but also bacteria, for problem solving in general. We focus especially on the concepts of swarm evolution, collective intelligence, stygmergic communication and self-organization. Before discussing these related insights, we will motivate the emergence of this discipline and thereafter present the major issue of modelling NP-hard problems using this technology. The field of swarm intelligence resides on the frontier between Multi-agents Systems (MAS) and Evolutionary Algorithms (EA). It has grown very rapidly the last decade and various swarm algorithms have been developed. We will study three methods in a chronological order of their publications. The first one, the Ant Colony Optimization (ACO) has known wide applications and use in industries. The Particle Swarm Optimization (PSO) will be introduced next, and the last one will be the Bat Algorithm (BA), recently proposed in the literature. The latter is stimulating a great interest from the artificial intelligence community. The other aspects common to the evolutionary algorithms such as parameters setting are also presented. At last, we will exhibit real applications from diverse domains using swarm intelligence to show its usefulness for the informatics industry as well as for the research area. To be more concrete, we will work on a mini-project consisting in applying Bat Algorithm for a robot movement in a 2D space with obstacles.
Keywords: collective intelligence, stygmergy, self-organization, Ants systems, Particle Swarm Optimization, Bat algorithms.
A glance at the course plan
Chapter 1. Introduction to Swarm Intelligence
Overview on Complex problems and NP-hardness Necessity of Collective intelligence for problem solving Stymergy and self-organization
A rich literature viewed through bee swarm intelligence
Chapter 2. Ant Colony Optimization (ACO)
From biology psychology to swarm algorithms
Different ACO approaches: AC, ACS, Elitist Ant, RBAS, BWAS and MMAS A simple application and its modelling with ACO
Chapter 3. Particle Swarm Optimization (PSO)
Philosophy of the approach PSO Algorithm
A simple application with PSO
Chapter 4. Bat Algorithm (BA)
Bat swarm and its specificities BA algorithm
A simple application of BA
Chapter 5. A mini-project: a robot movement in a 2D space with obstacles
Description of the project Modelling with Bat Algorithm Experimentations
Schedule
- January 25, 10:00 – 13:00, Seminar room, DISCo
- February 1, 10:00 – 13:00, Seminar room, DISCo
- February 4, 10:00 – 13:00, Seminar room, DISCo
- February 8, 10:00 – 13:00, Seminar room, DISCo
- February 18, 10:00 – 13:00, Meeting room of the III floor, DISCo
- February 22, 10:00 – 13:00, Seminar room, DISCo
- February 25, 10:00 – 13:00, Seminar room, DISCo