Evolutionary Fuzzy Systems: A Case Study For Intrusion Detection Systems

The so-called Evolutionary Fuzzy Systems consists of the application of evolutionary algorithms in the design process of fuzzy systems. Thanks to this hybridization, excellent abilities are provided to fuzzy systems in different work scenarios of data mining, such as standard classification, regression problems and association rule mining. The main reason of their success is the adaptation of their inner characteristics to any context. Among different areas of application, Evolutionary Fuzzy Systems have recently excelled in the area of Intrusion Detection Systems, yielding both accurate and interpretable models. To fully understand the nature and goodness of these type of models, we will introduce a full taxonomy on Evolutionary Fuzzy Systems. Then, we will overview a number of proposals from this research area that have been developed to address Intrusion Detection Systems. Finally, we will present a case study highlighting the good behaviour of Evolutionary Fuzzy Systems in this particular context.