WCSE

Predictive Analysis of Cloud Incidents

Abstract— With the widespread use of cloud computing, the number of cloud incidents involving outages, vulnerabilities, data loss, auto fails and hacks are constantly increasing. Although several prediction models have been proposed to forecast cloud incidents, such models do not consider trend, level, and seasonality components of cloud incidents. Using time series analysis, we create a predictive model for cloud incidents. Results show that the level of the series to be the best estimator of the prediction model and that time series model can be useful for prediction.

Index Terms— cloud incidents, prediction, time series, ARIMA

Yaman Roumani
Eastern Michigan University, Ypsilanti, US
Joseph K. Nwankpa
The University of Texas Rio Grande Valley, US
Yazan F. Roumani
Oakland University, US

[Download]

Cite: Yaman Roumani, Joseph K. Nwankpa, Yazan F. Roumani, "Predictive Analysis of Cloud Incidents," Proceedings of 2017 the 7th International Workshop on Computer Science and Engineering, pp. 448-453, Beijing, 25-27 June, 2017.