MltGNN: A Multi-level Temporal Graph Neural Network for Workload Prediction in Cloud

Key Points

• To develop a multi-level temporal graph neural network that accurately predicts workloads in cloud environments.
• Developed a multi-level temporal graph neural network architecture.
• Tested the model using various cloud workload datasets.
• Compared predictions with traditional methods to evaluate performance.
• Achieved higher accuracy in workload predictions compared to conventional models.
• Showed improved efficiency in resource allocation due to better predictions.