报告人：Soumik Sarkar, Ph.D., Assistant Professor, Department of Mechanical Engineering, Iowa State University

报告题目：Engineering Analytics with Machine Learning: Roles of supervised, semi-supervised and unsupervised models

报告时间：2018年7月4日（星期三）10：00

报告地点：李兆基科技大楼A459会议室

讲演者简介：

Dr. Soumik Sarkar received his B. Eng. Degree in Mechanical Engineering in 2006 from Jadavpur University, Kolkata, India. He received M.S. in Mechanical Engineering and M.A. in Mathematics in 2009 from Penn State University. Dr. Sarkar received his Ph.D. in Mechanical Engineering from Penn State in 2011. He joined the Department of Mechanical Engineering at Iowa State as an Assistant Professor in Fall 2014. Previously, he was with the Decision Support & Machine Intelligence group at the United Technologies Research Center for 3 years as a Senior Scientist. Dr. Sarkar’s research interests include Machine Learning, Sensor Fusion, Fault Diagnostics & Prognostics, Distributed Control and Complexity Analysis with applications to applications to complex Cyber-Physical Systems such as aerospace, energy & smart building systems, transportation, manufacturing and agriculture systems. He co‐authored more than 110 peer-reviewed publications including 45 journal papers and 5 book chapters. Dr. Sarkar is currently serving as an Associate Editor of Frontiers in Robotics and AI: Sensor Fusion and Machine Perception journal and an NVIDIA DLI University Ambassadorship. His research was supported by over $7M federal and private funding over the past four years. Dr. Sarkar is a recipient of the prestigious US NSF CISE Career Initiation Initiative (CRII) award in 2015 and the Young Investigator award from the US Air Force Office of Scientific Research (AFOSR) in 2017.

报告摘要：

Over the past few years, Machine Learning has experienced increasing popularity in solving difficult engineering problems ranging from design, manufacturing to system performance monitoring and control. Furthermore, with the advent of deep learning, the capability of handling high levels of system complexity and very large data sets has enhanced dramatically. This talk will discuss three recent success stories of Machine and Deep Learning for engineering analytics that are rather nontraditional in the context of computer science. First, I will share some recent supervised deep learning case studies in design optimization for microfluidic lab-on-chip devices and design for manufacturability for fast and democratized product design. I will then discuss how we used semi-supervised deep learning models for early detection of flame instability in combustion processes from hi-speed flame images in order to prevent catastrophic lean blow out in aircraft and other engines. Finally, I will conclude with a complex human-engineered system monitoring application that leverages unsupervised machine learning models.