Γενετικός Αλγόριθμος Ταξινόμησης Genetic AIRS
View/ Open
Subject
Μαθηματικά -- Πληροφορική ; MATLABAbstract
This thesis presents an hybrid classification method based on one hand on theories and principles of Artificial
Immune Systems and on the other hand on Genetic Algorithm techniques. The inspiration came from AIRS
algorithm(Artificial Immune Resource System), one of the most accurate classification algorithms, and from
Genetic Algorithm methods, so we developed one hybrid algorithm Genetic AIRS. Next, we investigated
the behavior of the algorithm created, with respect to the AIRS algorithm and present the advantages and
disadvatages. The main difference of these two algorithms beside the fact of different implementation is the
fact that AIRS executes topically and Genetic AIRS spherically. The first part of this thesis is the theoritical
base of the algorithm and the second part is the application on various data sets.