lifelonglab/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… See the full description on the dataset page: https://huggingface.co/datasets/lifelonglab/MCAD-CIC-3x1.