Unsupervised Real-Time Stream-Based Novelty Detection Technique an Approach in a Corporate Cloud
- 1 August 2018
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 166-170
Abstract
A highly loaded cloud application environment requires the highest stability and operability, generates large telemetry data streams. These are obvious and actual prerequisites to develop a workload shift detector for the failures prevention aim. Having studied the previous works, the authors developed an approach to the detection of changepoints based on the specific conditions of the streaming telemetry data. The simulation of data center workload has allowed us to generate telemetry data under specific workload, thus we can evaluate the performance of the detector under various conditions. The conducted experiment has shown the viability of the proposed approach as well as directions for further study and improvement.Keywords
This publication has 11 references indexed in Scilit:
- Manifold learning techniques for unsupervised anomaly detectionExpert Systems with Applications, 2018
- Unsupervised real-time anomaly detection for streaming dataNeurocomputing, 2017
- Scalable real-time classification of data streams with concept driftFuture Generation Computer Systems, 2017
- Ensemble learning for data stream analysis: A surveyInformation Fusion, 2017
- On Atomic Batch Executions in Stream ProcessingProcedia Computer Science, 2016
- The Entropy and PCA Based Anomaly Prediction in Data StreamsProcedia Computer Science, 2016
- Distributed Real-Time Anomaly Detection in Networked Industrial Sensing SystemsIEEE Transactions on Industrial Electronics, 2014
- Basic Singular Spectrum Analysis and forecasting with RComputational Statistics & Data Analysis, 2014
- An Anomaly Detection Approach Based on Isolation Forest Algorithm for Streaming Data using Sliding WindowIFAC Proceedings Volumes, 2013
- New ensemble methods for evolving data streams2009