Figure 1: From left to right, top to bottom, the training process of our GAN network is depicted. The generated data is plotted as a density chart throughout the training process, showing how the network learns to reflect the fidelity of the original data.
Generative Adversarial Networks (GAN) are proved effective for generating synthetic data. However, they fall behind when it comes to generating time-series data, which is observed in various real-world systems that deal with scheduling, load balancing, congestion control, etc. This work analyses how different GAN architectures, based on RNN, CNN, and transformers, can be used to generate time-series datasets with various characteristics. Throughout the experiments, the paper gives insights into which GAN architecture best fits each dataset type. All the reviewers acknowledged the importance of the problem and agreed on the usefulness of the study.
The complexity of computing along the cloud-to-edge continuum presents significant challenges to ICT operations and in particular reliable capacity planning and resource provisioning to meet unpredictable, fluctuating, and mobile demand. This chapter presents a high-level 2
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.