Figures & data
Figure 1. Illustration of the quadratic underestimating function . The objective function F is evaluated at the set of sample points
(illustrated by the dots).
![Figure 1. Illustration of the quadratic underestimating function F¯. The objective function F is evaluated at the set of sample points V¯ (illustrated by the dots).](/cms/asset/f587b35e-f9aa-45e5-a5f6-1f17f84a5fa7/geno_a_1495716_f0001_b.gif)
Figure 2. Illustration in of the nearest edge projection splitting strategy. The split is identified across the edge
–
of the minimum spanning tree corresponding to the designs
by projecting (illustrated by the dashed line) the solution to (Equation3
(3a)
(3a) ), in terms of
, onto the tree.
![Figure 2. Illustration in R2 of the nearest edge projection splitting strategy. The split is identified across the edge Zi(3)–Zi(5) of the minimum spanning tree corresponding to the designs Zi(Xis)={Zi(1),…,Zi(6)} by projecting (illustrated by the dashed line) the solution to (Equation3(3a) minimizev,λF¯(v1,…,vm),(3a) ), in terms of vi, onto the tree.](/cms/asset/fb0975bd-25ed-466a-9768-62afda110433/geno_a_1495716_f0002_b.gif)
Figure 5. Data profiles and performance profiles
for three variants of quadDS, the MATLAB GA and NOMAD applied to 120 instances of the sparse artificial problem with
.
![Figure 5. Data profiles da(β) and performance profiles ρa(α) for three variants of quadDS, the MATLAB GA and NOMAD applied to 120 instances of the sparse artificial problem with τ=0.1.](/cms/asset/11a41db6-7a6c-4980-a818-000e4a26ba8c/geno_a_1495716_f0005_b.gif)
Figure 6. Data profiles and performance profiles
for three variants of quadDS, the MATLAB GA and NOMAD applied to 120 instances of the fully artificial problem with
.
![Figure 6. Data profiles da(β) and performance profiles ρa(α) for three variants of quadDS, the MATLAB GA and NOMAD applied to 120 instances of the fully artificial problem with τ=0.1.](/cms/asset/0dcf9ead-a3e5-45b7-aed2-df4760a8385c/geno_a_1495716_f0006_b.gif)
Figure 7. Data profiles and performance profiles
for three variants of quadDS, the MATLAB GA and NOMAD applied to 120 instances of the beam problem with
.
![Figure 7. Data profiles da(β) and performance profiles ρa(α) for three variants of quadDS, the MATLAB GA and NOMAD applied to 120 instances of the beam problem with τ=0.1.](/cms/asset/9628e631-e4c3-462c-ac85-7cc60b3707d7/geno_a_1495716_f0007_b.gif)
Figure 8. Data profiles and performance profiles
for three variants of quadDS, the MATLAB GA and NOMAD applied to 120 instances of the tyres selection problem with
. Note that
for all values of β and
for all values of α, because NOMAD did not solve any of the instances of the tyres selection problem within the maximum allowed number of objective function evaluations.
![Figure 8. Data profiles da(β) and performance profiles ρa(α) for three variants of quadDS, the MATLAB GA and NOMAD applied to 120 instances of the tyres selection problem with τ=0.001. Note that dNOMAD(β)=0 for all values of β and ρNOMAD(α)=0.5 for all values of α, because NOMAD did not solve any of the instances of the tyres selection problem within the maximum allowed number of objective function evaluations.](/cms/asset/376230eb-965e-4e57-a1f7-086cf6a3d330/geno_a_1495716_f0008_b.gif)