Recall and precision control in anomaly detection

An approach to control the recall and precision in an anomaly detection system is presented using support vector machine (SVM). Genetic Algorithm (GA) is used to optimize the feature set and to train the model of SVM. The chromosome in GA consists of feature selection and training model. The fitness of chromosome is the (formulation) of recall and precision which are computed by the ξα-estimate method. A parameter is given to control the (recall) and precision in the formulation. The experiment results show that when the parameter increases, the (recall) increases and the precision decreases. It permits users to control the recall and precision by setting the value of the parameter.