Cyber Digital Twin with Deep Learning Model for Enterprise Products Management

Time series abnormalities might be signs of upcoming problems; thus, new computational anomaly detection techniques are needed for early warning systems and real-time system condition monitoring. Security and intrusion detection systems (IDS) are critical components of Internet of Things (IoT) devices. Current approaches are inadequate for handling complex data and unique intrusion detection systems (IDSs) in today’s network security platforms; deep learning techniques are needed. The Cybertwin-Enhanced self-attention mechanisms with long short-term memory anomaly detection (DL-Cyberwin-Enhanced SAM-LSTM-AD) model for business solutions that may achieve higher prediction accuracy for IoT devices is the main component of this suggested study. It is based on deep learning. To find the absolute error rate threshold of a new model, this model examines assaults against the Cybertwin-neural network. We looked at the CSE-CIC-IDS-2018 dataset to gauge the classifiers’ performance. Utilising the model's capacity to do time series analysis, this research combines the processed data into its suggested framework. These models show that the suggested model is feasible based on the high true positive rate (TPR) and low false positive rate (FPR) that were achieved. The model is evaluated using the test dataset using important metrics including F1-score, ROC-AUC, TPR, FPR, accuracy, and precision.