Professor Dr. Mian Muhammad Awais

(Lahore University of Management Sciences)

"Peircian Evolutionary Algorithm"

In the scientific community evolution has been equated with Darwinism. Whereas Darwin did indeed produce one of the first systematic arguments evidencing evolution, alternative arguments exist as well. Charles Sanders Peirce, an American philosopher and scientist, has put forth one such theory. The fundamental difference between Darwin's theory of evolution and Peirce's theory is the following. Darwin's theory is based on a dyadic logic that sees evolution being the result of interaction between two agents: blind chance and immutable mechanistic laws, including natural selection. Peirce's theory is based on a triadic logic that posits that evolution is the result of a dynamic interaction between three agencies: spontaneity, necessity and regularity. The theoretical level Peirce's theory is able to give a more comprehensive account of the evolutionary process in empirical reality, especially in light of the latest findings in molecular biology, epigentics and systems biology. Secondly, in experimental terms, we are making a maiden claim that Peirce's insights are useful enough to improve the existing Evolutionary Algorithms (EA) in computer science. Evolutionary Algorithms based on Peirce's theory of evolution solve some of the major noted problems in this area such as loss of diversity, stagnation, or premature convergence. In this regard we have provided a novel algorithm based on Peirce's theory of evolution, and tested it extensively against a benchmark set of mathematical problems of varying dimensions and complexity. Although we feel that there is still a long way to go, this research provides the preliminary ideas necessary to show the theoretical as well as empirical superiority of Peirce's framework.



Zeit: Donnerstag, 05.01.2012, 15.00 Uhr
Ort: Gebäude 48, Raum 680