resource allocation

STARIXNet: Multivariate and Multi-attribute Deep Learning Approach to Real-Time Resource Allocation in Cloud Platforms

The paper introduces STARIXNet, a lightweight neural network for resource allocation in cloud platforms, addressing the limitations of current univariate solutions that neglect risks of underestimation and delays. This deep learning approach captures spatio-temporal relationships and multiple attributes to guide intelligent microservices scaling decisions.