A survey of deep learning models used in proactive forecasting of cloud data center resources

In a cloud architecture, it is possible to make the most of cloud data center resources such as storage, networking, database services, content delivery networks, and other services. Cloud providers must have the ability to make accurate scaling decisions in advance to avoid overload by taking a proactive and predictive approach when designing cloud data center infrastructure. Saving resources is extremely important for reducing costs and ensuring sustainability. Therefore, we find that the proactive forecasting approach to anticipate patterns of resource use is one of the most important strategies for achieving efficient use of resources, as proactive forecasting includes predicting future results before they occur. One of the most important techniques used in proactive forecasting is deep learning because of its superiority in processing unstructured data and its ability to learn complex patterns. The aim of this research paper is a comprehensive survey of some deep learning models used in proactive prediction of cloud data center resources, such as convolutional neural networks (CNNs), recurrent neural networks (RNNs), long short-term memory networks (LSTMs), and hybrid models (CNN_LSTMS). This paper will present evidence from different scientific papers, offering a balanced perspective on the benefits of deep learning networks in proactive prediction.