anonymizeddb/

MCAD-CIC-3x1

	MCAD-CIC-3x1




	
	Dataset Summary

MCAD-CIC-3x1 is a multi-source continual anomaly detection benchmark scenario for network intrusion detection. It combines three CIC-family source datasets into a three-task continual-learning scenario:

cicids2017 cicids2018 cicunsw

Each task corresponds to one consolidated source dataset. The benchmark is designed to evaluate continual anomaly detection methods under cross-source distribution shift. The dataset contains 17,915,569 samples… See the full description on the dataset page: https://huggingface.co/datasets/anonymizeddb/MCAD-CIC-3x1.