The depth-design: An efficient generation of high dimensional computer experiments

Piffl, Martin; Stadlober, Ernst

doi:10.1016/j.jspi.2015.03.002

Cited by 3 publications

(2 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…5 DSDs are popular starting designs in the RSO literature 7 and recently Ref. [29] use DSDs as starting designs for screening in computer experiments.…”

Section: Stage 1: Screeningmentioning

confidence: 99%

Robust sequential experimental strategy for black‐box optimization with application to hyperparameter tuning

Sunder¹,

Albrecht²,

Nachtsheim³

2022

Quality & Reliability Eng

View full text Add to dashboard Cite

Global optimization of expensive, noisy black‐box functions is a problem of significant interest and is used extensively in applications such as hyperparameter optimization (HPO) of machine learning models. Response Surface Optimization (RSO) and Bayesian Optimization (BO) are two highly efficient model‐based experimentation strategies for optimizing expensive, noisy black‐box functions. The run‐size efficiency of both RSO and BO is critically dependent on the fitness of the surrogate model in approximating the true response function. In this regard, both RSO and BO deliberately choose initial designs that are optimal for estimating the assumed surrogate model and their sequential design strategy is aimed at further refining the surrogate model estimate. As a result, a design strategy that is optimal for one is not necessarily optimal for the other. In this study, we propose Compound‐RSO, a model‐robust, three‐stage experimental strategy for optimizing noisy black‐box functions with continuous experimental factors. Compound‐RSO is a hybrid strategy between RSO and BO which empirically estimates the complexity of the black‐box function and chooses the appropriate strategy between them. We propose robust designs that are highly efficient for estimating a second‐order polynomial model and that simultaneously maximize the power of the Lack‐of‐Fit test to determine the inadequacies of a second‐order approximation. Additionally, we propose a trust‐region‐based‐adaptive experimentation strategy for optimizing a second‐order response function. Through our simulation study we illustrate that the proposed Compound‐RSO strategy is more efficient than BO when the response function is second‐order, and performs comparably to BO when the response function is complex. Lastly, a case study on HPO of Deep Neural Networks (DDNs) in the context of visual quality inspections at a medical device manufacturer further illustrates the utility of the proposed Compound‐RSO strategy.

show abstract

“…5 DSDs are popular starting designs in the RSO literature 7 and recently Ref. [29] use DSDs as starting designs for screening in computer experiments.…”

Section: Stage 1: Screeningmentioning

confidence: 99%

Robust sequential experimental strategy for black‐box optimization with application to hyperparameter tuning

Sunder¹,

Albrecht²,

Nachtsheim³

2022

Quality & Reliability Eng

View full text Add to dashboard Cite

show abstract

“…In general, the approximate model needs a lot of samples. Orthogonal experiment [17][18][19], Latin hypercube design [20], and so forth can reduce the number of samples under the same approximate calculation accuracy.…”

Section: Introductionmentioning

confidence: 99%

Bending Angle Prediction Model Based on BPNN‐Spline in Air Bending Springback Process

Guo

Tang

2017

Mathematical Problems in Engineering

View full text Add to dashboard Cite

In order to rapidly and accurately predict the springback bending angle in V-die air bending process, a springback bending angle prediction model on the combination of error back propagation neural network and spline function (BPNN-Spline) is presented in this study. An orthogonal experimental sample set for training BPNN-Spline is obtained by finite element simulation. Through the analysis of network structure, the BPNN-Spline black box function of bending angle prediction is established, and the advantage of BPNN-Spline is discussed in comparison with traditional BPNN. The results show a close agreement with simulated and experimental results by application examples, which means that the BPNN-Spline model in this study has higher prediction accuracy and better applicable ability. Therefore, it could be adopted in a numerical control bending machine system.

show abstract

Inverse Determination of B340LA Material Parameters in Bending Springback Process by Dynamic Optimization Approach

Guo

Zhou

et al. 2021

Intelligent Robotics and Applications

View full text Add to dashboard Cite

The depth-design: An efficient generation of high dimensional computer experiments

Cited by 3 publications

References 17 publications

Robust sequential experimental strategy for black‐box optimization with application to hyperparameter tuning

Robust sequential experimental strategy for black‐box optimization with application to hyperparameter tuning

Bending Angle Prediction Model Based on BPNN‐Spline in Air Bending Springback Process

Inverse Determination of B340LA Material Parameters in Bending Springback Process by Dynamic Optimization Approach

Contact Info

Product

Resources

About