Automatic data featurization for enhanced proactive service auto-scaling: Boosting forecasting accuracy and mitigating oscillation

Keyphrases

• Auto-scaling 100%
• Cool-down 20%
• Dynamic Workload 20%
• Edge Computing 20%
• Edge Nodes 20%
• Extracting Features 20%
• Featurization 100%
• IoT Edge Computing 20%
• IoT System 20%
• Limited Resources 20%
• Long Short-term Memory 20%
• Machine Learning Techniques 20%
• Model Accuracy 20%
• Oscillation 100%
• Oscillation Phenomenon 20%
• Predicting the Future 20%
• Prediction Accuracy 100%
• Prediction Model 20%
• Prediction Speed 20%
• Proactive Auto-scaling 20%
• Proactive Services 100%
• Proactive Solution 20%
• Reactive Approach 20%
• Resource Utilization 20%
• Scalable Performance 20%
• Scaling Solutions 20%
• Service Deployment 20%
• System Workload 20%
• Time Series Data Analysis 20%
• Time-critical Applications 20%
• Widespread Adoption 20%

Computer Science

• Dynamic Workload 33%
• Edge Computing 33%
• Experimental Result 33%
• Extracted Feature 33%
• forecasting accuracy 100%
• Internet-Of-Things 66%
• Long Short-Term Memory Network 33%
• Machine Learning Technique 33%
• Model Accuracy 33%
• Prediction Accuracy 100%
• Prediction Model 33%
• Proactive Service 100%
• Resource Utilisation 33%
• Sensitive Application 33%
• Time Series Data 33%
• Time-Sensitive 33%
• Widespread Adoption 33%