Thursday July 28/3:15
Design and Analysis of Computer Experiments
The selection of inputs and analysis of output from execution of complex computer codes may be usefully treated from the perspective of statistical science. Existing knowledge and experience with experimental design is useful in efficiently selecting inputs for a computer code, whether the purpose be for exploring the code, predicting the output at other inputs, finding an extremum of a function, or tuning the code to an experimental or observational data base. In this minisymposium, the speakers will present research which connects such novel approaches with more classical numerical analytic and statistical ideas, and new applications to problems that seem otherwise intractable.
Organizer: Dennis D. Cox
- 3:15: Interpolation of Frequency Response Using a Single Model for Statistical Circuit Design.
Robert J. Buck, Ron Bates, and Henry P. Wynn, City University, London, United Kingdom
- 3:45: Designs for Computer Experiments and Quadrature.
Art B. Owen, Stanford University
- 4:15: Fractional Factorials and Bayesian Prediction.
Donald Ylvisaker, University of California, Los Angeles
- 4:45: Estimating Parameters in Complex Computer Models.
Dennis D. Cox, Organizer; Clifford E. Singer, University of Illinois, Urbana; and Jeong Soo Park, Chonnam National University, Korea