Event活動公告

2022.05.12(Thu.)「Quantum Machine Learning and cloud Computing」

Title-Quantum Machine Learning and cloud Computing
Date-2022/5/12
Time-13:00 – 14:00 (Taipei, UTC+8)
Register Link-https://forms.gle/9bV1iKcrLkV5yguh6

ONLINE Google Meet(http://meet.google.com/ykf-ovrv-qwj)

Speaker- Dr. Deepika Saxena (Department of Computer Applications, National Institute of Technology)
Abstract:
Cloud environments enabled with minimum upfront capital investment and maximum scalability features allow end users to expand and shrink their demand of resources dynamically over time. However, the fluctuations in the resource demands and pre-defined size of virtual machines (VMs), sharing of common physical machines among multiple users leads to resource wastage, excessive power consumption, increased security breaches and performance degradation. To address these pivotal and complex challenging issues, this presentation addresses three contributions including: (1) an Evolutionary Quantum Neural Network (EQNN) model towards prediction of dynamic and extensive range of cloud workloads, (2) an Online Predictive and Multi-objective Load Balancing (OP-MLB) Framework for an effective resource management, (3) an Online Secure inter-VM Communication Model (OSC-MC) for secure execution of sensitive workloads in shared computing environment. The EQNN model is an ingenious collaboration of computational efficiency of Quantum mechanics and adaptive machine learning capabilities of evolutionary neural networks. The OP-MLB framework develops and incorporates VM prediction, resource distribution and allocation with intended VM migration at unified platform, and allows interaction among them to optimize and tune together for overall performance improvement of cloud services. OSC-MC furnishes cybersecurity by identifying and terminating malicious VMs and inter-VM links prior to occurrence of security threats. All these works have been implemented and evaluated using wide variety of benchmark cloud workloads. The achieved results and and comparison with the state-of-the-art approaches validated the influential performance and potency of the proposed works.

Leave a Reply

Your email address will not be published. Required fields are marked *