Title: Towards an Intelligent Self-Learning Network for Science

Abstract:  Without human intervention, network automation promises configuration, management, testing, deployment, and network infrastructure operations. Although a self-driving network is still at its early stage, many time-consuming and complex network management tasks are being automated; thanks to the virtualization of network components, software-defined networking, advancements in artificial intelligence, and insights learned from data. In the first part of my talk, I will discuss my work at Berkeley Lab and how we have explored novel machine learning techniques to allow distributed high-speed network infrastructures to learn intelligently. I will then conclude by demonstrating how we applied some of the novel techniques in multidisciplinary collaborative research with the applied physics division at the Lab.

Speaker:  Bashir Mohammed, Scientific Data Division, Berkeley Lab

Bio: Bashir Mohammed is a postdoctoral research scientist working on the Deep Learning and Artificial Intelligence High-Performance Networks (DAPHNE) project at Berkeley Lab. His current research focuses on developing AI and machine-learning algorithms to optimally control high-speed distributed network resources, minimize network downtime and avoid network traffic congestion for critical exascale scientific workflows. Bashir received his MS in Control Systems from the University of Sheffield and a Ph.D. in Computer Science from the University of Bradford UK, before moving to the United States in 2019 to start his postdoctoral work. He interned with Rolls-Royce in Sheffield, where he developed a control and optimization algorithm to model gas turbine engines. In 2019 he was among six postdoctoral fellows from US National Labs invited to Washington DC to speak to the Capitol Hill audience about their research. Outside the Lab, Bashir enjoys playing basketball while pursuing his love of acting on stage in local theatres.

Presenter Website:  https://bashirmohd.github.io/
Department Link:  https://crd.lbl.gov/divisions/scidata/sdm/staff/bashir-mohammed/

Host:   Jose Castillo (CSRC) and the Sustainable Horizons Institute CRLC Virtual Seminar Series