Table 4.

Comparisons of current research for handling the concept of drift and model decay in IDS.

Reference	Research Scope	Major Contribution	Challenges/Limitations
[32]	Single classifier	It involves a single classification method.	Poor accuracy results.
[33]	Ensemble classifier for active state	It involves multiple classification active state methods.	More accurate and efficient.
[34]	Ensemble classifier for passive state	It involves multiple classification methods for passive state data.	More accurate.
[35]	Chunk-based data level	It involves actual attribute selection, employing continuously adaptive statistics analyses of input and probabilities of class labels.	It can work on specific data types.
[36]	Online learning based	It involves the online learning method.	Work on limited data.