Predictive Analytics in Cloud Computing: An ARIMA Model Study on Performance Metrics

Abstract

Predictive analytics is a key aspect of cloud computing as it helps organizations to anticipate future events and take proactive measures to prevent issues before they occur. In this research, the goal was to perform an ARIMA (AutoRegressive Integrated Moving Average) model to predict cloud performance using various performance metrics. The study utilized ten different performance metrics, such as Response Time, Resource Utilization, Availability, Error Rate, Memory Usage, CPU Utilization, Disk I/O, Network Bandwidth and others to model cloud performance. The aim was to investigate the potential of ARIMA models to predict cloud performance by analyzing the impact of these different performance metrics on the model's accuracy. The study also used four performance criteria, namely LogL (Log Likelihood), AIC (Akaike Information Criterion), BIC (Bayesian Information Criterion), and HQ (Hannan-Quinn Criterion) to evaluate the performance of the ARIMA models. The results of the study showed that the ARIMA model (2,0) and (0,2) had the lowest AIC and BIC values among all the models considered. This indicated that these models were the most suitable for predicting cloud performance, as they had the lowest information loss compared to the other models. The results of the study provided evidence that ARIMA models can effectively predict cloud performance. This research highlights the importance of predictive analytics in cloud computing and the potential for ARIMA models to predict cloud performance. The findings have implications for organizations that rely on cloud computing. However, more research is needed in this area, as the study was limited to only ten performance metrics, and more extensive research is needed to validate the findings and to determine the best approach to predict cloud performance.