Proactive Computer Network Monitoring based on Homogeneous LSTM Ensemble

Computer networks are getting more complex these days. A computer network failure can result in the loss of important data, disruption of network services and applications, and economic loss and threaten national security. Therefore, it is crucial to detect failures on time and diagnose their root cause, which is possible with the help of proactive computer network monitoring. The paper proposes a conceptual model of a system for proactive computer network monitoring. Proactive monitoring is based on predicting the network behavior. To achieve high prediction accuracy, we propose to use a homogeneous ensemble, which consists of a single base learning algorithm. Base learning LSTM models for an ensemble of deep neural networks were created using the bagging algorithm. We use the CICIDS2017 intrusion detection evaluation dataset to evaluate the proposed approach. Experimental results show that our method is an effective approach to improving the accuracy of anomaly prediction in computer networks.