|
Abstract:
|
This purpose of this talk is to discuss methodologies for the design and analysis of binary response computer validation studies. These types of processes arise primarily from survivability testing of military equipment. The main objective of these studies is to compare the output of a computer model with the output of physical trials to ensure the computer model is operating appropriately. This talk will discuss a range of modeling methodologies, with the bulk of the attention motivated by the computer experiments literature regarding the prediction of unknown functions. Some less strenuous techniques, from simple Generalized Linear Models to an approximate Gaussian Stochastic Process technique will also be discussed. Finally, the design of such experiments will be briefly discussed.
|
