CMStatistics 2023: Start Registration
View Submission - CMStatistics
B1467
Title: Designs for computer experiments from sequences with zero autocorrelation function Authors:  Omar Alhelali - RMIT university (Australia) [presenting]
Stella Stylianou - RMIT University (Australia)
Stelios Georgiou - RMIT University (Australia)
Abstract: Constructing designs for computer experiments has gained a significant focus from scientists since physical experiments are sometimes costly or time-consuming. An approach for generating designs for computer experiments is presented with numerous factors and symmetrical runs. Sequences with zero autocorrelation functions, like T-sequences, Base sequences, and others, have been used to obtain the appropriate designs for computer experiments. Encouragingly, this method can quickly transform these sequences into designs for computer experiments without the need for additional computer searches. The resulting designs have desirable properties such as symmetry in the runs and orthogonality of any even-order effect with any main effect. The derived designs are suitable for fitting a response surface with the full second-order model.