WAIRS: improving classification accuracy by weighting attributes in the AIRS classifier

Abstract

Abstract-AIRS (Artificial Immune Recognition System) has shown itself to be a competitive classifier. It has also proved to be the most popular immune inspired classifier. However, rather than AIRS being a classifier in its own right as previously described, we see AIRS more as a pre-processor to a KNN classifier. It is our view that by not explicitly classing it as such development of this algorithm has been rather held back. Seeing it as a pre-processor allows inspiration to be taken from the machine learning literature where such pre-processors are not uncommon. With this in mind, this paper takes a core feature of many such pre-processors, that of attribute weighting, and applies it to AIRS. The resultant algorithm called WAIRS (Weighted AIRS) uses a weighted distance function during all affinity evaluations. WAIRS is tested on 9 benchmark datasets and is found to outperform AIRS in the majority of cases.

RETRACTED ARTICLE: Diagnosing breast cancer with an improved artificial immune recognition system

“…Here, kNN is one of the components that have been added to other elements and procedures to make what is known as AIRS; therefore, it could also be interpreted as a hybrid algorithm of AIRS-KNN. Some researchers consider AIRS as pre-processor for KNN (Seeker and Freitas 2007).…”

Learning the affinity measure in real and artificial immune systems

“…Secker and Freitas [13] used the mutual information approach for attribute weighting, and showed the improvement of AIRS algorithm performance.…”

An interdisciplinary perspective on artificial immune systems

“…Other AIS that have been inspired by the adaptive immune mechanisms of B-cells are AIRS [113], a supervised learning algorithm (or more recently described as an instance creation algorithm [93], and the B-cell algorithm [70], an optimisation algorithm with a unique contiguous hypermutation operator. Like negative selection, clonal selection has proven to be very popular in the AIS community spawning a great deal of research with recent examples including [23,62,77].…”