The Application of the Relative Entropy Density Divergence in Intrusion Detection Models

How to choose the best model based on the original audit data in intrusion detection system(IDS)? In this work, we use the relative entropy density divergence as a measure of the IDS models. Through the data´s probability distribution dependence analysis, carries on the comparison to the different IDS models based on the original audit data. The model whose probability distribution conforms to the real probability distribution of the data is the better one. Thus according to the data set´s feature to select the IDS model.