The Dendritic Cell Algorithm (DCA) and its improved version: Deterministic Dendritic Cell Algorithm (dDCA) are essentially binary classification algorithms based on the behavior of Dendritic Cells (DCs) in the immune system. Both DCA and dDCA collect and process the data in form of signals, and produce output signal. The signals are divided in two types: danger and safe signals, and the output signal is determined by the values of the danger and safe signals. However, both DCA and dDCA suffer from data misclassification due to their sensitivity to data order. In this research we proposed a Fuzzy Deterministic Dendritic Cell Algorithm (FdDCA), which combines dDCA, fuzzy sets, and K-means clustering. The main objective of this research is to smooth the sharp boundaries between signals since we cannot always identify a clear boundary between the values of the signals. Our approach fuzzifies the signal values using linguistic variables, and a rule base is built to support fuzzy inference. The experimental results based on real data sets show that our approach shows a promising results compared to DCA and dDCA.
|Title of host publication||GECCO '16 Companion Proceedings of the 2016 on Genetic and Evolutionary Computation Conference Companion|
|Editors||Tobias Friedrich, Frank Neumann, Andrew M. Sutton|
|Number of pages||4|
|Publication status||Published - 20 Jul 2016|
- artificial immune systems
- deterministic Dendritic Cell algorithm
- fuzzy logic