Intrusion Detection Methods Using an Ensemble of Decision Trees

Intrusion detection corresponds to a set of techniques that are used to find attacks which damages the computers and network infrastructures. Intrusion detection is a classification problem. Therefore data mining techniques can be used to classify a given network connection to either a normal connection or an anomaly connection. To do this various classification models can be used. Among all decision tree classifiers have become very popular because of its simplicity interpretability and its performance. However decision tree classifiers are known to have high variance. Therefore it is said to be an unstable classifier. Along with these the conventional decision tree classifier does not perform well when noise vagueness and uncertainty present in the data. However to resolve the above issues this book proposes to use an ensemble of decision tree classifiers. To show the effectiveness of the proposed methods various intrusion detection data sets along with standard data sets are used.