A New Version of the Deterministic Dendritic Cell Algorithm Based on Numerical differential and immune response

Abstract

The efficiency of the promising deterministic dendritic cell algorithm (dDCA) in terms of anomaly detection has been demonstrated in large number of studies. However, the results of these studies also show that the signal calculation phase of dDCA depends on artificial experience, while the algorithm also lacks adaptability. To overcome these limitations, we propose a new approach for anomaly detection based on dDCA, in which numerical differential and immune responses are applied. The numerical differential is introduced to provide a new method of calculating the signal values that is independent of artificial experience. In the immune response method, moreover, the immune nonlinear model is adopted to adjust the weight signal matrix of dDCA. The experimental results show that the proposed algorithm has significant advantages over the compared immune anomaly detection approaches in terms of signal calculation and adaptability.