ABSTRACT
The main goal of hybridisation is the integration of the advantages of the latest information technologies. There are two approach combinations: the combined use of different types of immune algorithms, and algorithms with modified immune operators, and the combined use of several paradigms and soft computing algorithms: neural networks, fuzzy neural networks, evolutionary algorithms, artificial immune systems. In this chapter the advantage of the proposed procedure of the clonal negative selection on the test data is shown by modifying ways of presenting the data and changing the mechanisms of generation and elimination of detectors which optimally chooses the detector size for solving binary classification problems. A modified negative selection procedure that uses optimisation as well as an artificial immune network for the optimisation of detector parameters are developed. A distinctive feature of this procedure is a modification of the learning process, through which the number and location of detectors is implemented to adaptive selection settings.
