Abstract: With the continuous development of cloud computing technology, cloud computing resource load changes exhibits more and more complex features. For the workload prediction problem of cloud computing resources, the linear and nonlinear characteristics of resource workload time series in cloud computing environment are considered comprehensively. This paper proposes a combined prediction model based on auto-regressive integrated moving average (ARIMA) and long short-term memory (LSTM). The experiments are carried out to compare the proposal and the traditional load prediction algorithm on the public dataset. The experimental results show that the cloud computing resource combination prediction model has significantly higher prediction accuracy than other prediction models, which significantly reduces the real-time prediction error of resource workload in the cloud environment.

Key words: cloud computing, resource management, load forecasting,
long short-term memory, auto-regressive integrated moving average