Month: April 2016

Professor Christos Georgakis presented a distinguished lecture on data-driven modeling

WP_20160426_14_08_22_Pro

Dr. Christos Georgakis is a Professor of Chemical and Biological Engineering at Tufts University where he has also been the Bernard M. Gordon Senior Faculty Fellow in Systems Engineering. He described two generalizations of the classical design of experiments (DoE) methodology, the long-standing data-driven modeling methodology of choice. The first generalization enables the design of experiments with time-varying inputs, called Design of Dynamic Experiments (DoDE). The second generalization enables the development of a dynamic response surface model (DRSM) when time-resolved measurements are available. He discussed how both advances are able to contribute significantly to the modeling, optimization, and understanding of processes for which a knowledge-driven model is not easily at hand. He also argued that such approaches can be widely used in developing reduced-size meta-models, for online use in existing processes.

To view more details or a video of the lecture, please visit http://utc-iase.uconn.edu/events/lec-lib/dll2016/.

Dr. Quan Long gives a seminar on efficient Bayesian optimal experimental design for physical models

IMG_0947

Dr. Quan Long from United Technologies Research Center provided an overview of his recent research on Efficient Bayesian Optimal Experimental Design for Physical Models. Optimal experimental design is the key to improve data quality in engineering. Its application on real problems lags behind mainly due to the involved computational costs. Dr. Long has developed a series of methods to accelerate the computations of the utility function (expected information gain) under rigorous error control. Specifically, he has extended the applicable domain of Laplace methods from the asymptotic posterior Gaussianity, to where the shape of the posterior is characterized by noninformative manifolds. While Laplace methods require a concentration of measure, multi-level Monte Carlo method can be used to efficiently compute the nested integral of the expected information gain with a reduction of the computational complexity, even when the randomness of data dominants the shape of the posterior distribution. The developed methodologies have been applied to various engineering problems, e.g., impedance tomography, seismic source inversion and parameter inference of combustion kinetics.

To view more details or a video of the seminar, please visit http://utc-iase.uconn.edu/events/sem-lib/sl2016/.