A generalized research study on distance measures, learning algorithms and datasets

Intrusion detection system is widely focused on improving the accuracy, time and space constraints by using various measures and techniques. Though many techniques are available to detect the known attacks, the research focus is on detecting the unknown attacks. Detecting attacks require using distance functions. The purpose of distance functions is to find the distance between known sample and new sample incoming into the network. Classifiers are used to classify incoming sample into either a safe or attack sample. This paper makes a survey from the point of view of distance functions, classifiers, datasets used to detect intrusions.