• Communications for Statistical Applications and Methods
• /
• v.24 no.4
• /
• pp.383-396
• /
• 2017

The model in our approach assumes that computer responses are a realization of a Gaussian processes superimposed on a regression model called a Gaussian process regression model (GPRM). Selecting a subset of variables or building a good reduced model in classical regression is an important process to identify variables influential to responses and for further analysis such as prediction or classification. One reason to select some variables in the prediction aspect is to prevent the over-fitting or under-fitting to data. The same reasoning and approach can be applicable to GPRM. However, only a few works on the variable selection in GPRM were done. In this paper, we propose a new algorithm to build a good prediction model among some GPRMs. It is a post-work of the algorithm that includes the Welch method suggested by previous researchers. The proposed algorithms select some non-zero regression coefficients (${\beta}^{\prime}s$) using forward and backward methods along with the Lasso guided approach. During this process, the fixed were covariance parameters (${\theta}^{\prime}s$) that were pre-selected by the Welch algorithm. We illustrated the superiority of our proposed models over the Welch method and non-selection models using four test functions and one real data example. Future extensions are also discussed.

• ETRI Journal
• /
• v.43 no.1
• /
• pp.74-81
• /
• 2021

The mixture model is a very powerful and flexible tool in clustering analysis. Based on the Dirichlet process and parsimonious Gaussian distribution, we propose a new nonparametric mixture framework for solving challenging clustering problems. Meanwhile, the inference of the model depends on the efficient online variational Bayesian approach, which enhances the information exchange between the whole and the part to a certain extent and applies to scalable datasets. The experiments on the scene database indicate that the novel clustering framework, when combined with a convolutional neural network for feature extraction, has meaningful advantages over other models.

• Journal of the Korean Society for Nondestructive Testing
• /
• v.28 no.1
• /
• pp.54-58
• /
• 2008

Time reversal (TR) of body waves in fluids and isotropic solids has been used in many applications including ultrasonic NDE. However, the study of the TR method for anisotropic materials is not well established. In this paper, the full reconstruction of the input signal is investigated for anisotropic media using an analytical formulation, called a modular Gaussian beam (MGB) model. The time reversal operation of this model in the frequency domain is done by taking the complex conjugate of the Gaussian amplitude and phase received at the TR mirror position. A narrowband reference signal having a particular frequency and number of cycles is then multiplied and the whole signal is inverse Fourier transformed. The original input signal is seen to be fully restored by the TR process of MGB model and this model can be more generalized to simulate the spatial and temporal focusing effects due to TR process in anisotropic materials.

• The Korean Journal of Applied Statistics
• /
• v.26 no.3
• /
• pp.389-399
• /
• 2013

This article introduces Gaussian process regression and shows its application with time-course microarray gene expression data. Gene screening for yeast cell cycle microarray expression data is accomplished with a ratio of log marginal likelihood that uses Gaussian process regression with a squared exponential covariance kernel function. Gaussian process regression fitting with each gene is done and shown with the nine top ranking genes. With the screened data the Gaussian model-based clustering is done and its silhouette values are calculated for cluster validity.

• ETRI Journal
• /
• v.32 no.5
• /
• pp.795-800
• /
• 2010

Embedding a large vocabulary speech recognition system in mobile devices requires a reduced acoustic model obtained by eliminating redundant model parameters. In conventional optimization methods based on the minimum description length (MDL) criterion, a binary Gaussian tree is built at each state of a hidden Markov model by iteratively finding and merging similar mixture components. An optimal subset of the tree nodes is then selected to generate a downsized acoustic model. To obtain a better binary Gaussian tree by improving the process of finding the most similar Gaussian components, this paper proposes a new distance measure that exploits the difference in likelihood values for cases before and after two components are combined. The mixture weight of Gaussian components is also introduced in the component merging step. Experimental results show that the proposed method outperforms MDL-based optimization using either a Kullback-Leibler (KL) divergence or weighted KL divergence measure. The proposed method could also reduce the acoustic model size by 50% with less than a 1.5% increase in error rate compared to a baseline system.

• Nuclear Engineering and Technology
• /
• v.52 no.8
• /
• pp.1807-1816
• /
• 2020

In this work, a Gaussian Process (Kriging) approach is proposed to provide efficient dose mapping for complex radiation fields using limited number of responses. Given a few response measurements (or simulation data points), the proposed approach can help the analyst in completing a map of the radiation dose field with a 95% confidence interval, efficiently. Two case studies are used to validate the proposed approach. The First case study is based on experimental dose measurements to build the dose map in a radiation field induced by a D-D neutron generator. The second, is a simulation case study where the proposed approach is used to mimic Monte Carlo dose predictions in the radiation field using a limited number of MCNP simulations. Given the low computational cost of constructing Gaussian Process (GP) models, results indicate that the GP model can reasonably map the dose in the radiation field given a limited number of data measurements. Both case studies are performed on the nuclear engineering radiation laboratories at the University of Sharjah.

• Communications of the Korean Mathematical Society
• /
• v.25 no.2
• /
• pp.291-301
• /
• 2010

This paper consists of two main parts. Firstly, we introduce an evolving binomial process from a binomial stock model and consider various types of limiting behavior of the logarithm of the evolving binomial process. Among others we find that the logarithm of the binomial process converges weakly to a Gaussian process. Secondly, we provide new approaches for proving the limit theorems for an integral process motivated by the evolving binomial process. We provide a new proof for the uniform strong LLN for the integral process. We also provide a simple proof of the functional CLT by using a restriction of Bernstein inequality and a restricted chaining argument. We apply the functional CLT to derive the LIL for the IID random variables from that for Gaussian.

• Wind and Structures
• /
• v.35 no.6
• /
• pp.419-430
• /
• 2022

In wind-resistant designs, wind velocity is assumed to be a Gaussian process; however, local complex topography may result in strong non-Gaussian wind features. This study investigates the non-Gaussian wind features over complex terrain under atmospheric turbulent boundary layers by the large eddy simulation (LES) model, and the turbulent inlet of LES is generated by the consistent discretizing random flow generation (CDRFG) method. The performance of LES is validated by two different complex terrains in Changsha and Mianyang, China, and the results are compared with wind tunnel tests and onsite measurements, respectively. Furthermore, the non-Gaussian parameters, such as skewness, kurtosis, probability curves, and gust factors, are analyzed in-depth. The results show that the LES method is in good agreement with both mean and turbulent wind fields from wind tunnel tests and onsite measurements. Wind fields in complex terrain mostly exhibit a left-skewed Gaussian process, and it changes from a softening Gaussian process to a hardening Gaussian process as the height increases. A reduction in the gust factors of about 2.0%-15.0% can be found by taking into account the non-Gaussian features, except for a 4.4% increase near the ground in steep terrain. This study can provide a reference for the assessment of extreme wind loads on structures in complex terrain.

• Nuclear Engineering and Technology
• /
• v.48 no.3
• /
• pp.684-701
• /
• 2016

The Gaussian process model (GPM) is a flexible surrogate model that can be used for nonparametric regression for multivariate problems. A unique feature of the GPM is that a prediction variance is automatically provided with the regression function. In this paper, we estimate the safety margin of a nuclear power plant by performing regression on the output of best-estimate simulations of a large-break loss-of-coolant accident with sampling of safety system configuration, sequence timing, technical specifications, and thermal hydraulic parameter uncertainties. The key aspect of our approach is that the GPM regression is only performed on the dominant input variables, the safety injection flow rate and the delay time for AC powered pumps to start representing sequence timing uncertainty, providing a predictive model for the peak clad temperature during a reflood phase. Other uncertainties are interpreted as contributors to the measurement noise of the code output and are implicitly treated in the GPM in the noise variance term, providing local uncertainty bounds for the peak clad temperature. We discuss the applicability of the foregoing method to reduce the use of conservative assumptions in best estimate plus uncertainty (BEPU) and Level 1 probabilistic safety assessment (PSA) success criteria definitions while dealing with a large number of uncertainties.

• Geomechanics and Engineering
• /
• v.28 no.1
• /
• pp.65-75
• /
• 2022

Time and cost of construction are key factors in decision-making during a tunnel project's planning and design phase. Estimations of time and cost of tunnel construction projects are subject to significant uncertainties caused by uncertain geotechnical and geological conditions. The Gaussian Process Regression (GPR) technique for predicting ground condition and construction time and cost of mountain tunnel projects is used in this work. The GPR model is trained with data from past mountain tunnel projects. The model is applied to a case study in which the predicted time and cost of tunnel construction using the GPR model are compared with the actual construction time and cost for model validation and reducing the uncertainty for the future projects. In addition, the results obtained from the GPR have been compared with to other models of artificial neural network (ANN) and support vector regression (SVR) that the GPR model provides more accurate results.