Immune-inspired Algorithm for Anomaly Detection

Recent Advances in Artificial Immune Systems: Models and Applications

“…Based on the combinations of the received signals, two terminal differentiation states, namely, mature or semi-mature, are generated during the maturation of dendritic cells. The mature dendritic cells exhibit the following behavior: collection of antigen ceases; expression of co-stimulatory molecules and chemical messengers (cytokines); migration from the tissue to a lymphatic organ; and presenting antigen to T lymphocytes [79]. Mature dendritic cells have an activating effect while semi-mature dendritic cells have a suppressive effect.…”

Conserved Self Pattern Recognition Algorithm

“…• The pre-defined anomalies are required to establish the PAMPs in recent work [10]. However, only normal data are used in CSPRA during the training phase and the special APC detector gains the capability of recognizing the conserved self pattern in the antigen object from extracting the useful knowledge from the Self.…”

“…1, are pre-defined bacterial signatures, causing the maturation of APCs through the expression of mature cytokines [10]. Each high-dimension antigen (or self protein) in our system is regarded as the information vector (or object).…”

“…The recent work [10] used the functionality of Dendritic Cell (DC) with Danger Theory as inspiration for developing the algorithm for anomaly detection. The author imported the concept of PAMPs in PRRs model to Danger Theory and regarded it as one of the danger signals within the tissue to determine the function that the DC will perform when the body is under attack.…”

“…Our work extensively mimics the PRRs model but PAMPs are built based on normal behaviors. PAMPs in [10] are arbitrary, depending on application domain. In our algorithm, PAMPs are well-defined mandatory components in the training stage based on the inputted normal data in the system.…”