An alternative formulation of the artificial immune recognition system learning algorithm
Μια εναλλακτική διαμόρφωση του αλγόριθμου μάθησης AIRS
KeywordsClassification ; Artificial Immune Recognition System ; AIRS ; Machine learning ; Artificial immune systems
The Artificial Immune Recognition System (AIRS) is an immune-inspired supervised learning algorithm that has been shown to perform competitively on some common data sets. The purpose of this thesis is the presentation of an alternative formulation of the Artificial Immune Recognition System, followed by a comparative study with emphasis on classification accuracy, data reduction capability and algorithmic efficiency in order to evaluate the performance difference between the proposed version of AIRS and the original AIRS classifier. The comparison suggests that the proposed formulation holds significant performance advantages over the original AIRS algorithm and further exploration of the main functionality of the algorithm could identify and address deficiencies that leave AIRS lacking in future research.