Facebook Twitter Instagram YouTube

Using Statistical and Machine-Learning Techniques for Understanding and Improving Complex Physical and Chemical Models

Event Details

Event Dates: 

Friday, April 27, 2012 - 10:00am

Speaker Name(s): 

Mike Davis

Speaker Affiliation(s): 

Chemical Sciences Division, Argonne National Laboratory
Seminar Type/Subject

Event Details & Abstract: 

This talk will discuss applications of techniques developed in the statistics and machine learning communities to problems in the physical sciences. The main focus will be applications of these techniques to chemical-kinetic modeling, but it will also present applications in nanophotonics and engine modeling. It will demonstrate how these approaches can be used for chemical-kinetic model improvement, optimization of the performance of nanophotonic devices, and the evaluation of the importance of individual chemical reactions on engine performance. In the last case, it will show how a detailed understanding of the role of hindered rotation in chemical reactions is important for properly modeling ignition in engine simulations.The talk reports on worked performed in collaboration with many researchers, particularly:   Dingyu Zhou, Rex Skodje, and Lawrence Harding; Wei Liu, RaghuSivaramakrishnan, Sibendu Som, and Douglas Longman; Ryan Miller and Stephen Gray.

JILA follows the six University nodes' policies for ensuring harassment-free environments. For more detailed information regarding the University of Colorado policies, please read the Discrimination and Harassment Policy and Procedures.