Meetings of the Biodiversity & Ecology Journal Club Department of Biological Sciences, The National University of Singapore Featuring seminar announcements from Ecotax

Meetings and seminars about biology by the staff, students and visitors of NUS - Archives

List of Categories : admin * bejc * dbs * ecotax *

Tuesday, 8th November 2005: Department Of Biological Sciences Special Seminar

Category : dbs

Evolutionary Computing: Potential Applications in Environmental and Biodiversity Research

by Vladan Babovic
Associate Professor
Department of Civil Engineering &
Tropical Marine Science Institute
National University of Singapore

Tuesday, 8th November 2005: 10.00am
Seminar Room 3
Blk S2, Level 2, Department of Biological Sciences
The National University of Singapore
Science Drive 4

Visitors may park at Carpark 10

View the map

Host: A/P Peter Ng

About the Talk
Evolutionary algorithms (EAs) are engines simulating grossly simplified processes occurring in nature and implemented in artificial media Ñ such as a computer. Evolutionary algorithms effectively provide an alternative approach to problem solving - where solutions of the problem are evolved rather than the problems being solved directly. Today the family of evolutionary algorithms is divided into four main streams: Evolution Strategies, Evolutionary Programming, Genetic Algorithms and Genetic Programming.

Although different and intended for different purposes, all EAs share a common conceptual base. In principle, an initial population of individuals is created in a computer and allowed to evolve using the principles of inheritance (so that the offspring resembles the parents), variability (the process of offspring creation is not perfect - some mutations occur) and selection (more fit individuals are allowed to reproduce more often and less fit less often so that their ÔgenealogicalÕ trees disappear in time).

In genetic programming (GP) the evolutionary force is directed towards the creation of models that take a symbolic form. In this evolutionary paradigm, evolving entities are presented with a collection of data and the evolutionary process is expected to result in a closed-form symbolic expression describing the data. GP iteratively applies variation and selection on a population of evolving parse trees representing symbolic

expressions. Standard variation operators in genetic programming are subtree mutation (replace a randomly chosen subtree with a randomly generated subtree) and subtree crossover (replace a randomly chosen subtree from a formula with a randomly choÐsen subtree from another formula). The types of functions used in this tree structure are user-defined. This means that they can be algebraic operators, such as sin, log, +, -, etc., but they can also take the form of if-then-else rules, making use of logical operators such as OR, AND, etc.

This talk introduces the genetic programming in the context of aquatic environment. The main idea is the one of using the creative power of genetic programming to discover and optimize the structure of a model, whereas the evaluation of these models is carried out in symbolic math environments. Several practical applications are presented in order to demonstrate effectiveness and usefulness of this approach. The experimental results show that these kinds of algorithms introduce numerous advantages over most available modeling methods.

ALL ARE WELCOME!

Posted at 12:26AM SGT | permalink | , .

Meetings of the Biodiversity & Ecology Journal Club.
Archives - About BEJC - Email webmaster

Guidelines for hosts/proposers

NTU Natural Sciences Academic Group (NSAG) - Dept Seminars - Graduate Seminars