Convergent Design Sequences, for Sufficiently Regular Optimality Criteria
Atwood, Corwin L.
Ann. Statist., Tome 4 (1976) no. 1, p. 1124-1138 / Harvested from Project Euclid
For an optimality criterion function $\Phi$ and a design $\xi_n$, approximate $\Phi(\mathbf{M}(\xi))$ near $\xi_n$ by a quadratic Taylor expansion, let $\xi_n + \eta$ minimize this approximation, and let $\xi_{n+1} = \xi_n + \alpha\eta$, with $\alpha$ minimizing $\Phi(\mathbf{M}(\xi_{n+1}))$. If $\Phi$ satisfies regularity conditions, including strict convexity, possession of three continuous derivatives, and finiteness only for nonsingular $\mathbf{M}$, then $\mathbf{M}(\xi_n)$ converges to the optimal value for both the Federov steepest descent sequence and the above quadratic sequence, with the quadratic sequence having a faster asymptotic convergence rate. Methods are discussed for collapsing clusters of design points during the iterative process. In a simple example with $D$-optimality, the two methods are comparable. In a more complicated example the quadratic method is far superior.
Publié le : 1976-11-14
Classification:  Optimal experimental designs,  $D$-optimality,  $L$-optimality,  $\phi$-optimality,  steepest descent,  generalized Newton method,  62K05,  65B99
@article{1176343647,
     author = {Atwood, Corwin L.},
     title = {Convergent Design Sequences, for Sufficiently Regular Optimality Criteria},
     journal = {Ann. Statist.},
     volume = {4},
     number = {1},
     year = {1976},
     pages = { 1124-1138},
     language = {en},
     url = {http://dml.mathdoc.fr/item/1176343647}
}
Atwood, Corwin L. Convergent Design Sequences, for Sufficiently Regular Optimality Criteria. Ann. Statist., Tome 4 (1976) no. 1, pp.  1124-1138. http://gdmltest.u-ga.fr/item/1176343647/