What Machine Learning can do for Science and Engineering?
2 pm, Tuesday, May 23, 2017; Room 304, Institute for Cyber-systems & Control, Zhejiang University（浙江大学玉泉校区智能系统与控制研究所304教室，联系人：许超，13706711953）
Speaker: Guang Lin（林光）, PhD
Associate Professor, Department of Mathematics & School of Mechanical Engineering Purdue University
Machine learning has attracted a lot attention recently. In this talk, I will use several case studies to demonstrate what machine learning can do for Science and Engineering.
It is observed that birds, bats, insects, and fish can routinely harness unsteady fluid phenomena to improve their propulsive efficiency, maximize thrust and lift, and increase maneuverability. In this talk, I will demonstrate how to use machine learning strategy to characterize the time-varying fluid flows with very limited sensor information in modern engineering, for instance, biological propulsion and bio-inspired engineering design.
In addition, I will also present how to use machine learning techniques to employ very limited satellite data or sensor information to improve the climate model predictive capability or identify the contaminant source locations. Particularly, an adaptive importance sampling technique will be introduced to utilize machine learning method to capture multimodal distribution using a mixture of Gaussian distribution.
Ebola disease has been taking thousands of people’s life. It is critical to develop effective strategy to prevent Ebola’s outbreak. I will present how we can use machine learning techniques to develop more accurate Ebola model using limited data. In addition, we employ this accurate Ebola model to develop and test several different strategies to studies its effectiveness in preventing Ebola’s outbreak.
Guang Lin got his bachelor in Zhejiang university in 1997 and PhD from division of applied mathematics at Brown university in 2007. Now he is an associate professor in both department of mathematics and school of mechanical engineering at Purdue University.
Guang Lin received NSF faculty early career development award in recognition of his work on uncertainty quantification and big data analysis in smart grid and other complex interconnected systems. Guang Lin has developed advanced optimization algorithms to calibrate complex global and regional climate models. For this work, he received a Ronald L. Brodzinski Award for Early Career Exception Achievement in 2012. Guang Lin also received 2010ASCR Leadership Computing Challenge (ALCC) award in recognition of his work in analyzing big climate data using extreme-scale supercomputers. Guang Lin has also received Outstanding Performance Award at Pacific Northwest National Laboratory in 2010, and Ostrach Fellowship at Brown University in Fall 2005.