batch.adsa: Calculates reallocated batch size or new input location for...

In mludkov/mlOSP: Machine Learning and Regression Monte Carlo Algorithms for Optimal Stopping

Calculates reallocated batch size or new input location for Adaptive Design with Sequential Allocation

Description

ADSA for Adaptive Batching

Usage

batch.adsa(fit, r_seq, xtest, xt_dens, x_new, r0, nugget, method)

Arguments

fit	GP/TP fit
r_seq	batch size vector for existing inputs
xtest	testing points to compare reallocation and adding a new inputs
xt_dens	density of xtest
x_new	new input location selected by the EI criteria
r0	total number of new simulations
nugget	the noise variance to compute the ALC factor
method	km or trainkm or hetgp or homtp

Value

a list containing

• xoptim: new design input (NULL if re-allocation is chosen)

• roptim: added replications (a scalar r0 if new input chosen, a vector containing the re-allocation amounts otherwise

Author(s)

Xiong Lyu

References

M. Ludkovski, X. Lyu (2020+) Adaptive Batching for Gaussian Process Surrogates with Application in Noisy Level Set Estimation, <http://arxiv.org/abs/2003.08579>

See Also

osp.seq.batch.design

mludkov/mlOSP documentation built on April 29, 2023, 7:56 p.m.