Back to the program.
Observer Paradigm: A new way to look at a landscape.
Christophe Philemotte
cphilemo [at] iridia [dot] ulb [dot] ac [dot] be


Bigger the search space is, more time is needed for investigating the search space. For instance, the Simple Genetic Algorithm (SGA) is not totally suited to traverse very large landscapes, especially deceptive ones. The paradigm proposed here aims at improving the way a given metaheuristic goes through the search space. A second search process is then engaged: an ``observer'' is defined as each possible encoding that aims at reducing the search space and adequacy of one observer is dependent of how it is benificial for the metaheuristic process. A general overviw of our approach is proposed, and some explanations are given about the effect of an ``observers'' on a landscape. The SGA as metaheuristic and the Shuffled Hierarchical IF-and-only-iF (SHIFF) as problem is chosen for illustrating these explanations.


Representation, Simple Genetic Algorithm, HIFF Problem, Intrinsic Emergence


  1. Christophe Philemotte and Hugues Bersini. (2006) How An Optimal Observer can Smooth a Landscape. In Proceedings of the 2006 IEEE Congress on Evolutionary Computation. IEEE. In press.
  2. Christophe Philemotte and Hugues Bersini. (2006) How an Optimal Observer can Collapse the Search Space,. In Proceedings of the 2006 conference on Genetic and evolutionary computation. ACM. In press.