Towards Automated and Efficient Machine Learning as a Service (MLaaS)

Speaker: Yan Feng, University of Nevada, Reno

Abstract
Driven by the sustained advances of Machine Learning (ML) and fast evolving of cloud computing, the past a few years have witnessed a surging demand of Machine Learning as a Service (MLaaS). MLaaS is widely supported by the leading cloud platforms to facilitate ML model design, model training, and model inference serving in an automated and scalable manner. In this talk, Dr. Yan will present the recent progress from his team on MLaaS, including improve automated ML model design with topology-aware Neural Architecture Search (NAS); develop an efficient training job deployment framework driven by a novel cost-ware Bayesian optimization approach; and tackle the dual challenges of Service Level Objective (SLO) compliance and cost effectiveness in inference model serving by designing a scalable and adaptive serving system. Dr. Yan will also introduce how these MLaaS techniques help wildfire smoke forecasting in a cross-discretionary NSF BIGDATA project led by Dr. Yan.

Bio
Feng Yan is an Assistant Professor in the Computer Science and Engineering Department at University of Nevada, Reno (UNR). He is the director of the Intelligent Data and Systems Lab (IDS Lab) at UNR. Dr. Yan received both the M.S. degree and Ph.D. degree in Computer Science from the College of William and Mary, and worked at Microsoft Research and HP Labs. Dr. Yan’s recent research focus includes machine learning, cloud/edge/high performance computing, storage, as well as cross- disciplinary topics such as big data driven wildfire forecasting, AI strengthened material science and engineering. Dr. Yan and his team are actively publishing at the most prestigious venues in machine learning area (e.g., NIPS, KDD, AAAI, etc.) and computer system area (e.g., USENIX ATC, SC, HPDC, FAST, EuroSys, etc.). Dr. Yan and his students are recipients of the Best Student Paper Award of IEEE CLOUD 2018 and the Best Paper Award of CLOUD 2019. Dr. Yan serves as Social Media Chair of ACM SIGMETRICS.