Negative Selection: How to Generate Detectors

The immune system is a remarkable and complex natural system, which has been shown to be of interest to computer scientists and engineers alike. This paper reports an on-going investigation into the usefulness of the negative selection metaphor for immune inspired fault tolerance. Various procedures to generate detectors for the negative selection algorithm are reviewed and compared in terms of time and space complexity for the production of competent detectors. A new algorithm has been identified and implemented. Experimentation was undertaken, and an analysis is presented on the effectiveness of the various algorithms. The outcome of this empirical analysis reveals that trade-offs have to be made in the choice of algorithm based on the time and space complexities, as well as the detection rate.