2014
DOI: 10.1016/j.csda.2013.02.021
|View full text |Cite
|
Sign up to set email alerts
|

Computing efficient exact designs of experiments using integer quadratic programming

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
2
1
1
1

Citation Types

1
27
0

Year Published

2015
2015
2020
2020

Publication Types

Select...
5
1

Relationship

2
4

Authors

Journals

citations
Cited by 21 publications
(28 citation statements)
references
References 34 publications
1
27
0
Order By: Relevance
“…We also remark that, using a standard computer, AQuA based on the IQP solver (unlike the specific MICQP formulation of AQuA) cannot be applied to design spaces of size larger than a few thousands, because of the quadratic memory requirements. That is, here we again demonstrated the advantage of the proposed conic AQuA approach over the approach of AQuA from Harman and Filová [13], not only over methods directly based on a MISOCP formulation of the problem as in Sagnol and Harman [34].…”
Section: D-and I-optimal Subsampling Of a Dataset Under An Upper Conssupporting
confidence: 59%
See 2 more Smart Citations
“…We also remark that, using a standard computer, AQuA based on the IQP solver (unlike the specific MICQP formulation of AQuA) cannot be applied to design spaces of size larger than a few thousands, because of the quadratic memory requirements. That is, here we again demonstrated the advantage of the proposed conic AQuA approach over the approach of AQuA from Harman and Filová [13], not only over methods directly based on a MISOCP formulation of the problem as in Sagnol and Harman [34].…”
Section: D-and I-optimal Subsampling Of a Dataset Under An Upper Conssupporting
confidence: 59%
“…Harman and Filová [13] proposed a substantially different approach to the use of an optimal AD for ED construction, which overcomes many disadvantages of ER and similar methods. In particular, it does not depend on the choice of the optimal AD if the AD is not unique, it is not restricted to the support of the optimal AD, and the resulting EDs are usually significantly more efficient than the EDs computed by ER.…”
Section: Let ξ Ementioning
confidence: 99%
See 1 more Smart Citation
“…However, the performance of these methods is not always good; the search for optimal designs is often restricted to low-dimensional models. New developments help to overcome these limitations (see Yu 2011;Sagnol 2011;Harman and Filová 2014).…”
Section: Computing Optimal Designsmentioning
confidence: 98%
“…The algorithm proposed here is in the general class of the multiplicative algorithms (Silvery, et al, 1978), which shares the simplicity and monotonic convergence property of class of the multiplicative algorithm. One of the most attractive properties of the proposed algorithm is that the convergence rate of the algorithm does not depend on the number of design points N compared to some existing algorithms such as the coordinate-exchange algorithm (Meyer and Nachtsheim, 1995) and the randomized exchange algorithm (Harman and Filová, 2014).…”
Section: Algorithm For D-optimal Designsmentioning
confidence: 99%