Author:
M. P. Iwundu,G. O. Agadaga
Abstract
New non-sequential third-order response surface designs are proposed with good optical properties. They are suitable as one-stage experimental designs for use in modeling third-order effects. The new designs are presented for cuboidal region in k dimensions and the technique employed in the construction of the non-sequential designs on the cuboidal region is flexible for use in regions that may be non-cuboidal. The new non-sequential designs lay importance on the use of axial points and two or three other blocks of points selected from a discrete design region such that the design is non-singular. For a continuous design region, uniform grids are formed over the entire design region. Five grid levels are utilized in this work thus resulting in 5^k grid points from which blocks of points are selected to form the desired non-sequential designs. The goodness of the designs is assessed via optimality and efficiency criteria and the new designs possess good optimality properties and are very high by G-efficiency.
Publisher
African - British Journals
Reference37 articles.
1. Alaeddini, A., Yang, K., Murat, A. (2013b). ASRSM: A Sequential Experimental Design for Response Surface Optimization. Quality and reliability Engineering International, 29(2): 241-258.
2. Anjaneyulu, G.V.S.R., Dattatreya Rao, A.V. and Narasimham, V.L. (2001). Third Order Slope
3. Anjaneyulu, G.V.S.R., Dattatreya Rao, A.V. and Narasimham, V.L. (1994 - 1995) Construction of third order slope rotatable designs , Gujarat Statistical Review Vols.21 &22,49-53
4. Arshad, H. M., Akhtar, M. and Gilmour, S. G. (2012). Augmented Box-Behnken designs for fitting third order response surfaces. Communications in Statistics – Theory and Methods 41(23): 4225-4239.
5. Baker, F. D. and Bargmann, R. E. (1985). Orthogonal Central Composite Designs of the Third Order in the Evaluation of Sensitivity and Plant Growth Simulation Models. Journal of the American Statistical Association, Vol. 80. No. 391391, pp. 574-579.