Fig 1.
The framework of the proposed MICAM.
Table 1.
The methods for data anomaly detection.
Table 2.
1–9 scale method.
Table 3.
The value condition of RI.
Table 4.
The effects of integrity, accuracy, consistency, diversity, and balance in the power system scenario.
Fig 2.
The distribution of CICIDS2017 dataset.
Fig 3.
The evaluation results of indicators for CICIDS2017 dataset.
Fig 4.
The weight coefficients of evaluation indicators.
Fig 5.
The distribution of intrusion detection data type.
Fig 6.
The evaluation results of indicators for intrusion detection data.
Table 5.
The precision and recall of different AI models with p < 0.05.
Fig 7.
The comparisons of F1-score of different AI models with p < 0.05.