Sequential modeling of a low noise amplifier with neural networks and active learning

Gorissen, Dirk; Tommasi, Luciano De; Crombecq, Karel; Dhaene, Tom

doi:10.1007/s00521-008-0223-1

Cited by 70 publications

(50 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…All these experiments were carried out using the SUMO Toolbox research platform [3,4]. This freely available Matlab toolbox, designed for adaptive surrogate modelling and sampling, has excellent extensibility, making it possible for the user to add, customize and replace any component of the sampling and modelling process.…”

Section: Resultsmentioning

confidence: 99%

Generating Sequential Space-Filling Designs Using Genetic Algorithms and Monte Carlo Methods

Crombecq

Dhaene

2010

Lecture Notes in Computer Science

Self Cite

View full text Add to dashboard Cite

show abstract

Section: Resultsmentioning

confidence: 99%

Generating Sequential Space-Filling Designs Using Genetic Algorithms and Monte Carlo Methods

Crombecq

Dhaene

2010

Lecture Notes in Computer Science

Self Cite

View full text Add to dashboard Cite

show abstract

“…However, in this work it is shown how this type of kernels applied to non-stationary problems can be troublesome in combination with sequential design. Consider, for example, Figure 2a, showing a sequentially designed dataset of a Low Noise Amplifier (LNA) (Gorissen et al 2009) and the resulting optimized standard GP model, using a stationary SE kernel. The sequential dataset has a much higher sampling density in the irregular regions of the response surface (e.g.…”

Section: Gaussian Processesmentioning

confidence: 99%

“…As a second application we visit the Low Noise Amplifier (LNA) modeling problem (Gorissen et al 2009). An LNA is an electronic amplifier used to amplify very weak signals.…”

Section: Low Noise Amplifiermentioning

confidence: 99%

Deep Gaussian Process metamodeling of sequentially sampled non-stationary response surfaces

Dutordoir¹,

Knudde²,

Herten³

et al. 2017

2017 Winter Simulation Conference (WSC)

Self Cite

View full text Add to dashboard Cite

Simulations are often used for the design of complex systems as they allow to explore the design space without the need to build several prototypes. Over the years, the simulation accuracy, as well as the associated computational cost has increased significantly, limiting the overall number of simulations during the design process. Therefore, metamodeling aims to approximate the simulation response with a cheap-to-evaluate mathematical approximation, learned from a limited set of simulator evaluations. Kernel-based methods using stationary kernels are nowadays wildly used. However, using stationary kernels for non-stationary responses can be inappropriate and result in poor models when combined with sequential design. We present the application of a novel kernel-based technique, known as Deep Gaussian Processes, which is better able to cope with these difficulties. We evaluate the method for non-stationary regression on a series of real-world problems, showing that it outperforms the standard Gaussian Processes with stationary kernels.

show abstract

“…A full description of the LNA problem can be found in [17]. The chosen model type for this problem is artificial neural networks, trained with Levenberg-Marquard backpropagation with Bayesian regularization (300 epochs).…”

Section: Case 3: Low-noise Amplifiermentioning

confidence: 99%

A Fuzzy Hybrid Sequential Design Strategy for Global Surrogate Modeling of High-Dimensional Computer Experiments

Herten¹,

Couckuyt²,

Deschrijver³

et al. 2015

SIAM J. Sci. Comput.

View full text Add to dashboard Cite

Abstract. Complex real-world systems can accurately be modeled by simulations. Evaluating high-fidelity simulators can take several days, making them impractical for use in optimization, design space exploration and analysis. Often, these simulators are approximated by relatively simple math known as a surrogate model. The data points to construct this model are simulator evaluations meaning the choice of these points is crucial: each additional data point can be very expensive in terms of computing time. Sequential design strategies offer a huge advantage over one-shot experimental design because information gathered from previous data points can be used in the process of determining new data points. Previously, LOLA-Voronoi was presented as a hybrid sequential design method which balances exploration and exploitation: the former involves selecting data points in unexplored regions of the design space, while the latter suggests adding data points in interesting regions which were previously discovered. Although this approach is very successful in terms of the required number of data points to build an accurate surrogate model, it is computationally intensive. This paper presents a new approach to the exploitation component of the algorithm based on Fuzzy logic. The new approach has the same desirable properties of the old method but is less complex, especially when applied to high-dimensional problems. Experiments on several test problems show the new approach is a lot faster, without losing robustness or requiring additional samples to obtain similar model accuracy.

show abstract

Sequential modeling of a low noise amplifier with neural networks and active learning

Cited by 70 publications

References 17 publications

Generating Sequential Space-Filling Designs Using Genetic Algorithms and Monte Carlo Methods

Generating Sequential Space-Filling Designs Using Genetic Algorithms and Monte Carlo Methods

Deep Gaussian Process metamodeling of sequentially sampled non-stationary response surfaces

A Fuzzy Hybrid Sequential Design Strategy for Global Surrogate Modeling of High-Dimensional Computer Experiments

Contact Info

Product

Resources

About