… This work aims to improve the resource contention problem for concurrent deep learning …
-wide resource management. The learning tasks considered are implemented as containerized …

… the uncertainty in real-life resource management and deal with the bursty or dramatic
decrease on workloads, we analyze the violation of SLAs after resource scheduling, and regard it …

… Input: Job pending queue groups: G, compaction threshold: N, parallelism for job j: pj, minimum
number of GPUs for job j: minj, resource manager that manages the free GPUs in cluster: …

… we presented a Deep Reinforcement Learning agent for cloud elasticity problems, called …
algorithmic techniques in both deep learning and cloud resource management areas. Our …

… efficient autonomic and elastic resource management approaches in the … resource
management, such as provisioning, allocation, scheduling, and monitoring, and also the elasticity …

… Lyra is a GPU cluster scheduler that exploits capacity loaning with elastic job scheduling. It
runs on top of a cluster resource manager such as YARN [51] and Kubernetes [23] to execute …

… resources utilizing resources effectively to achieve high-performance data analytics becomes
desirable. Although cluster resource management … machine learning (ML) / deep learning (…

… deep reinforcement learning (DRL) techniques in this work to obtain a resource management
… • We propose a deep reinforcement learning based approach utilizing LSTM model and …

… • Slice-aware resource management … the management and orchestration of future networks
proposed by ETSI. Then we discuss the application of AI in the context of resource elasticity …

… learning. ElasticFlow provides a serverless interface with two distinct features: (𝑖) users specify
only the deep neural network … level, and manual resource management for deep learning …