Event Detail

Event Type: 
Applied Mathematics and Computation Seminar
Friday, April 22, 2016 - 12:00 to 12:45
GLK 113

Speaker Info


We consider a data set resulting from experiments on a prestressed concrete girder. Given a priori engineering requirements, we derive an underlying model with spatially dependent elastic properties based on energy considerations. Least squares estimation techniques to calibrate a family of models with experimental data are considered. The possibility of nonunique estimators is also discussed. We then introduce a probabilistic inversion approach focusing on the ability of models to capture the information contained in the data set.