Efficient global optimization algorithm assisted by multiple surrogate techniques
Citations
421 citations
351 citations
330 citations
Additional excerpts
...Another two ways of using multiple surrogates in sequential sampling might be through 1) blind kriging [104,189], which can be seen as an ensemble, and 2) the use the pool of surrogates to providemultiple points per cycle of the EGO algorithm, such as in [190]....
[...]
295 citations
195 citations
Cites background from "Efficient global optimization algor..."
...Jones (2001) compares the performance of kriging-based surrogate models to quadratic non-interpolating models for global-search optimization, while Viana et al. (2013) develop approaches using multiple surrogate predictions to locate promising new sampling points within a box-constrained region....
[...]
References
24,053 citations
13,736 citations
Additional excerpts
..., radial basis neural networks [26,27], linear Shepard interpolation [28,29], and support vector regression [30, 31]....
[...]
10,696 citations
Additional excerpts
..., radial basis neural networks [26,27], linear Shepard interpolation [28,29], and support vector regression [30, 31]....
[...]
8,328 citations
6,914 citations
"Efficient global optimization algor..." refers background or methods in this paper
...Thus, the expected improvement E I (x) is the expectation of I (x) (derivation found in [7])...
[...]
...As defined by [7], the improvement at a point x is I (x) = max (yP BS − Y (x), 0) , (4)...
[...]
...For example, the efficient global optimization (EGO) algorithm [7] models the objective function as a random variable....
[...]