System of Covariance Data Evaluation at China Nuclear Data Centre-COVAC

R. R. Xu,T. J. Liu,J. S. Zhang,Z. J. Sun 한국물리학회 2011 THE JOURNAL OF THE KOREAN PHYSICAL SOCIETY Vol.59 No.23

To determine the uncertainty of neutron cross sections for modern nucleardata application, a new evaluation system, COVAC is being developed at ChinaNuclear Data Centre (CNDC). In COVAC, experimental data were firstlypre-analyzed and handled via processing tools. The sensitivities oftheoretical model parameters were analyzed by a special designedsubsystemSEMAW which offers a platform to obtain parameter sensitivities fordifferent nuclear reaction codes. In addition the generalized least-squaresmethod (GLS) and Bayesian method were both contained in COVAC to generatethe recommended covariance The experimental and theoretical uncertaininformation can be included in the covariance by each of the two methods.The outputs are formatted in ENDF-6. So far, COVAC is suitable to generatethe covariance files for structure and fission nuclei in the fast neutronenergy range, and n+^(48)Ti was taken to illuminate the wholeprocedure.

A transductive least squares support vector machine with the difference convex algorithm

Shim, Jooyong,Seok, Kyungha The Korean Data and Information Science Society 2014 한국데이터정보과학회지 Vol.25 No.2

Unlabeled examples are easier and less expensive to obtain than labeled examples. Semisupervised approaches are used to utilize such examples in an eort to boost the predictive performance. This paper proposes a novel semisupervised classication method named transductive least squares support vector machine (TLS-SVM), which is based on the least squares support vector machine. The proposed method utilizes the dierence convex algorithm to derive nonconvex minimization solutions for the TLS-SVM. A generalized cross validation method is also developed to choose the hyperparameters that aect the performance of the TLS-SVM. The experimental results conrm the successful performance of the proposed TLS-SVM.

A transductive least squares support vector machine with the difference convex algorithm

심주용,석경하 한국데이터정보과학회 2014 한국데이터정보과학회지 Vol.25 No.2

Unlabeled examples are easier and less expensive to obtain than labeled examples. Semisupervised approaches are used to utilize such examples in an effort to boost the predictive performance. This paper proposes a novel semisupervised classification method named transductive least squares support vector machine (TLS-SVM), whichis based on the least squares support vector machine. The proposed method utilizesthe difference convex algorithm to derive nonconvex minimization solutions for the TLS-SVM. A generalized cross validation method is also developed to choose the hyperparameters that affect the performance of the TLS-SVM. The experimental results confirm the successful performance of the proposed TLS-SVM.

A transductive least squares support vector machine with the difference convex algorithm

Jooyong Shim,Kyungha Seok 한국데이터정보과학회 2014 한국데이터정보과학회지 Vol.25 No.2

Unlabeled examples are easier and less expensive to obtain than labeled exam-ples. Semisupervised approaches are used to utilize such exam-ples in an effort to boost the predictive performance. This paper proposes a novel semisupervised classification method named transductive least squares support vector machine (TLS-SVM), which is based on the least squares support vector machine. The proposed method utilizes the difference convex algorithm to derive nonconvex minimization solutions for the TLS-SVM. A generalized cross validation method is also developed to choose the hy-perparameters that affect the performance of the TLS-SVM. The experimental results confirm the successful performance of the proposed TLS-SVM.

Deep multiple kernel least squares support vector regression using PSO

Jooyong Shim,Insuk Sohn,Kyungha Seok 한국데이터정보과학회 2019 한국데이터정보과학회지 Vol.30 No.3

In this paper, we propose a deep multiple kernel least squares support vector regression (DMK-LSSVR) using particle swarm optimization (PSO). Unlike multilayer neural networks (MNN), each LSSVR in the DMK-LSSVR is trained to minimize the penalized objective function. Therefore, the learning of DMK-LSSVR is completely different from that of MNN that minimizes only the final objective function. In DMK-LSSVR the grid search using GCV function is used to find optimal values of hyperparameters of each LSSVR, which has a disadvantage that takes a lot of computational time. And the back propagation algorithm is used for the optimal values of weights and biases, which has a weakness to results in local minima. In DMK-LSSVR which utilizes PSO (DMK-LSSVR-PSO), we find the optimal values of hyperparameters of LSSVRs, weights and biases using PSO in one process. Using PSO, the only needed on hyperparameters are the lower and upper bound, and estimating weights and biases results in global minimizers. Numerical studies show that DMK-LSSVR-PSO has advantages over DMK-LSSVR and other machine learning models that use back propagation algorithm and grid search for regression problems.

Indirect Measurement of Brain Activation using fNIRS

M. Ahmad Kamran,Keum-Shik Hong 제어로봇시스템학회 2013 제어로봇시스템학회 국제학술대회 논문집 Vol.2013 No.10

Functional near infrared spectroscopy (fNIRS) is an emerging non-invasive brain imaging technique based on estimation of blood chromosomes. The change in the concentration of oxy-hemoglobin (HbO) and de-oxy hemoglobin (HbR) gives indirect measure of brain activation. fNIRS measures the change in HbO and HbR depending on the absorption of near infrared (NIR) light. It uses the NIR light of two wave lengths, 760 nm and 830 nm. NIRS is a newly developed neuro imaging modality with high temporal resolution. NIRS systems have great potential for brain computer interface due to its low cost, portability and safety. The scalp remains intact throughout the experiment. In this study we present a method for estimation of brain activation by using fNIRS data collected during arithmetic task. The general linear model has been used in this study with predicted hemodynamic response function, its derivatives and physiological noises as regressors. The normalized least mean square (NLMS) algorithm has been used for estimation of activity strength parameters in the model recursively. A one way t-test has been performed for detection of brain activation at different channels.

EXTENSION OF FACTORING LIKELIHOOD APPROACH TO NON-MONOTONE MISSING DATA

Kim, Jae-Kwang The Korean Statistical Society 2004 Journal of the Korean Statistical Society Vol.33 No.4

We address the problem of parameter estimation in multivariate distributions under ignorable non-monotone missing data. The factoring likelihood method for monotone missing data, termed by Rubin (1974), is extended to a more general case of non-monotone missing data. The proposed method is algebraically equivalent to the Newton-Raphson method for the observed likelihood, but avoids the burden of computing the first and the second partial derivatives of the observed likelihood. Instead, the maximum likelihood estimates and their information matrices for each partition of the data set are computed separately and combined naturally using the generalized least squares method.

Extension of Factoring Likelihood Approach to Non-monotone Missing Data

김재광 한국통계학회 2004 Journal of the Korean Statistical Society Vol.33 No.4

We address the problem of parameter estimation in multivariate distributions under ignorable non-monotone missing data. The factoring likelihood method for monotone missing data, termed by Rubin (1974), is extended to a more general case of non-monotone missing data. The proposed method is algebraically equivalent to the Newton-Raphson method for the observed likelihood, but avoids the burden of computing the rst and the second partial derivatives of the observed likelihood. Instead, the maximum likelihood estimates and their information matrices for each partition of the data set are computed separately and combined naturally using the generalized least squares method.

The factoring likelihood method for non-monotone missing data

Kim, Jae Kwang,Shin, Dong Wan 한국통계학회 2012 Journal of the Korean Statistical Society Vol.41 No.3

We address the problem of parameter estimation in multivariate distributions under ignorable non-monotone missing data. The factoring likelihood method for monotone missing data, termed by Rubin (1974), is applied to a more general case of non-monotone missing data. The proposed method is asymptotically equivalent to the Fisher scoring method from the observed likelihood, but avoids the burden of computing the first and second partial derivatives of the observed likelihood. Instead, the maximum likelihood estimates and their information matrices for each partition of the data set are computed separately and combined naturally using the generalized least squares method. A numerical example is presented to illustrate the method.

The factoring likelihood method for non-monotone missing data

김재광,신동완 한국통계학회 2012 Journal of the Korean Statistical Society Vol.41 No.3

We address the problem of parameter estimation in multivariate distributions under ignorable non-monotone missing data. The factoring likelihood method for monotone missing data, termed by Rubin (1974), is applied to a more general case of nonmonotone missing data. The proposed method is asymptotically equivalent to the Fisher scoring method from the observed likelihood, but avoids the burden of computing the first and second partial derivatives of the observed likelihood. Instead, the maximum likelihood estimates and their information matrices for each partition of the data set are computed separately and combined naturally using the generalized least squares method. A numerical example is presented to illustrate the method.