Choose a Section
Help
Local links

MA 798C
Numerical methods for model calibration
Goals of the course: This course is an introduction to model calibration and the optimization
technology used for model calibration. We will cover the following topics
 Review of probability concepts: random variables, probability distribution, law of large numbers, multivariate distribution, covariance matrix, Bayesian inference, examples
 Review of linear least squares: theory, computational methods, implementation, examples
 Kalman filtering: derivation, implementation, examples
 Computing gradients and Jacobians: sensitivities, adjoints, automatic differentiation
 Nonlinear least squares: theory, computational methods, implementation, examples
 Monte Carlo methods: pseudo random numbers, Monte Carlo integration, variance reduction, implementation, examples
 Optimization without derivatives: NelderMeade, implicit filtering
 Ensemble Kalman filtering: theory, computational methods, implementation, examples
 Examples
Grade: will be determined by the completion of a individual project (to be discussed with the instructors) to be presented in class.
Notes: the instructors will make their notes available
Office hours: TBA
Contact info:
 Pierre Gremaud, Dept. of Math, NCSU, Raleigh,
NC 276958205, phone: 9195153085, fax: 9195151636, email:
gremaud@ncsu.edu
 Tim Kelley, Dept. of Math, NCSU, Raleigh,
NC 276958205, phone: 9195157163, fax: 9195151636, email:
tim_kelley@ncsu.edu
