Immune Network for Classifying Heterogeneous Data

Abstract

In the previous AIS research, most of the AIS classifiers use clonal selection and require the data to be in numerical or categorical data types prior to processing. These classifiers ignore the network feature of the immune system that is suitable for classification. Furthermore, the transformation of data into any other specific types from their original form can degrade the originality of the data and consume more space and pre processing time. This paper introduces resource limited immune network model with hybrid affinity measurement for classifying heterogeneous data in its original types. The model is able to process the data with the types as represented in the database. The paper shows comparisons between the model and the selected existing immune algorithms that also uses the same set of data and parameters. The experimental results show that the immune network model produces a better accuracy rate with shorter classifier on most of the heterogeneous data from University of California, Irvive (UCI) machine learning repository (MLR).