Robust population designs for longitudinal linear regression model with a random intercept
- 71 Downloads
In this paper, optimal population designs for linear regression model with a random intercept for longitudinal data are considered. The design space is assumed to be a set of equally spaced time points. Taking the sampling scheme for each subject as a multidimensional point in the space of admissible sampling sequence, we determine the optimal number and allocation of sampling times in order to estimate the fixed effects model as accurately as possible. To make comparisons between different designs in a meaningful manner, we take experimental costs into account when defining the D-optimal design criterion function. We take a Bayesian method to overcome the uncertainty of the parameters in the design criterion to derive Bayesian optimal population designs. For complicated cases, we propose a hybrid algorithm to find optimal designs. Meanwhile, we apply the Equivalence Theorem to check the global optimality of these designs.
KeywordsOptimal design Mixed effects models Equivalence theorem Particle swarm optimization
This work was partially supported by NSFC Grant (11301332, 11471216), China.
- Debusho LK (2004) Optimal designs for linear mixed models. PhD thesis, South Africa, University of KwaZulu-NatalGoogle Scholar
- Duchateau L, Janssen P, Rowlands GJ (1998) Linear mixed models. An introduction with applications in veterinary research. International Livestock Research Institute, NairobiGoogle Scholar
- Eberhart RC, Kennedy J (1995) A new optimizer using particle swarm theory. In: Proceedings of the sixth international symposium on micro machine and human science(Nagoya, Japan). IEEE Service Center, Piscataway, pp 39–43Google Scholar
- Kennedy J, Eberhart RC (1997) A discrete binary version of the particle swarm algorithm. In: IEEE international conference on in systems, man, and cybernetics, vol 5. IEEE, pp 4104–4108Google Scholar
- Wong WK, Chen RB, Huang CC, Wang W (2015) A modified particle swarm optimization technique for finding optimal designs for mixture models. PLoS ONE. https://doi.org/10.1371/journal.pone.0124720