HIGHLIGHTS
SUMMARY
Previously proposed machine_learning methods do not identify how anomalous behaviour caused by cyber-attacks propagates and damages connected devices in ICSs. Existing machine_learning-based solutions provide local anomaly detection, and they do not consider other correlated events that can be the source of another attack. It can be noticed that deep learning methods such as LSTM outperform any other machine_learning-based anomaly detection approach as the scale of data increases. The challenge of applying machine_learning algorithms on a high dimensional and low-value density dataset is that highly correlated features may skew the results . . .
If you want to have access to all the content you need to log in!
Thanks :)
If you don't have an account, you can create one here.