• Communications for Statistical Applications and Methods
• /
• v.16 no.6
• /
• pp.925-936
• /
• 2009

There very many approaches to impute missing values in the iid. case. However, it is hardly found the imputation techniques in the Markov random field(MRF) case. In this paper, we show that the imputation under MRF is just to impute by fitting the normal mixture model(NMM) under several practical assumptions. Our multivariate normal mixture model based approaches under MRF is applied to impute the missing pixel values of 3-variate (R, G, B) color image, providing a technique to smooth the imputed values.

• Speech Sciences
• /
• v.14 no.1
• /
• pp.151-162
• /
• 2007

This study focuses on the classification of pathological voice using GMM (Gaussian Mixture Model) and compares the results to the previous work which was done by ANN (Artificial Neural Network). Speech data from normal people and patients were collected, then diagnosed and classified into two different categories. Six characteristic parameters (Jitter, Shimmer, NHR, SPI, APQ and RAP) were chosen. Then the classification method based on the artificial neural network and Gaussian mixture method was employed to discriminate the data into normal and pathological speech. The GMM method attained 98.4% average correct classification rate with training data and 95.2% average correct classification rate with test data. The different mixture number (3 to 15) of GMM was used in order to obtain an optimal condition for classification. We also compared the average classification rate based on GMM, ANN and HMM. The proper number of mixtures on Gaussian model needs to be investigated in our future work.

• Journal of the Korean Statistical Society
• /
• v.28 no.2
• /
• pp.167-182
• /
• 1999

This article is concerned with the selection of subsets of predictor variables to be included in building the binary response probit regression model. It is based on a Bayesian approach, intended to propose and develop a procedure that uses probabilistic considerations for selecting promising subsets. This procedure reformulates the probit regression setup in a hierarchical normal mixture model by introducing a set of hyperparameters that will be used to identify subset choices. The appropriate posterior probability of each subset of predictor variables is obtained through the Gibbs sampler, which samples indirectly from the multinomial posterior distribution on the set of possible subset choices. Thus, in this procedure, the most promising subset of predictors can be identified as the one with highest posterior probability. To highlight the merit of this procedure a couple of illustrative numerical examples are given.

• The Korean Journal of Applied Statistics
• /
• v.29 no.3
• /
• pp.513-523
• /
• 2016

Fitting a mixture of multivariate skew normal distribution (MSNMix) with multiple skewness parameter vectors via EM algorithm often requires a highly expensive computational cost to calculate the moments and probabilities of multivariate truncated normal distribution in E-step. Subsequently, it is common to fit an asymmetric data set with MSNMix with a simple skewness parameter vector since it allows us to compute them in E-step in an univariate manner that guarantees a cheap computational cost. However, the adaptation of a simple skewness parameter is unrealistic in many situations. This paper proposes an approximate estimation for the MSNMix with multiple skewness parameter vectors that also allows us to treat them in an univariate manner. We additionally provide some experiments to show its effectiveness.

• Journal of the Korean Statistical Society
• /
• v.29 no.4
• /
• pp.407-422
• /
• 2000

This article is concerned with the selecting predictor variables to be included in building a class of binary response t-link regression models where both probit and logistic regression models can e approximately taken as members of the class. It is based on a modification of the stochastic search variable selection method(SSVS), intended to propose and develop a Bayesian procedure that used probabilistic considerations for selecting promising subsets of predictor variables. The procedure reformulates the binary response t-link regression setup in a hierarchical truncated normal mixture model by introducing a set of hyperparameters that will be used to identify subset choices. In this setup, the most promising subset of predictors can be identified as that with highest posterior probability in the marginal posterior distribution of the hyperparameters. To highlight the merit of the procedure, an illustrative numerical example is given.

• The Korean Journal of Applied Statistics
• /
• v.25 no.4
• /
• pp.651-659
• /
• 2012

In this paper we developed a statistical model for circular data with noises. In this case, model fitting by single circular model has a lack-of-fit problem. To overcome this problem, we consider some mixture models that include circular uniform distribution and apply an EM algorithm to estimate the parameters. Both von Mises and Wrapped skew normal distributions are considered in this paper. Simulation studies are executed to assess the suggested EM algorithms. Finally, we applied the suggested method to fit 2008 EHFRS(Epidemic Hemorrhagic Fever with Renal Syndrome) data provided by the KCDC(Korea Centers for Disease Control and Prevention).

• The Korean Journal of Applied Statistics
• /
• v.16 no.2
• /
• pp.305-319
• /
• 2003

This paper provides a non-Bayesian method based on the expanded EM algorithm for segmenting the magnetic resonance images degraded by bias field. For the images with the intensity as a pixel value, many segmentation methods often fail to segment it because of the bias field(with low frequency) as well as noise(with high frequency). Our contextual approach is appropriately designed by using normal mixture model incorporated with Markov random field for noise-corrective segmentation and by using the penalized likelihood to estimate bias field for efficient bias filed-correction.

• Journal of the Korean Data and Information Science Society
• /
• v.28 no.2
• /
• pp.251-260
• /
• 2017

The option pricing of Black와 Scholes (1973) and Merton (1973) has been widely reported to fail to reflect the time varying volatility of financial time series in many real applications. For example, Duan (1995) proposed GARCH option pricing method through Monte Carlo simulation. However, financial time series is known to follow a fat-tailed and leptokurtic probability distribution, which is not explained by Duan (1995). In this paper, in order to overcome such defects, we proposed the option pricing method based on GARCH models with normal mixture errors. According to the analysis of KOSPI200 option price data, the option pricing based on GARCH models with normal mixture errors outperformed the option pricing based on GARCH models with normal errors in the unstable period with high volatility.

• MALSORI
• /
• no.66
• /
• pp.61-72
• /
• 2008

This paper proposes a method to improve pathological and normal voice classification performance by combining multiple features such as auditory-based and higher-order features. Their performances are measured by Gaussian mixture models (GMMs) and linear discriminant analysis (LDA). The combination of multiple features proposed by the frame-based LDA method is shown to be an effective method for pathological and normal voice classification, with a 87.0% classification rate. This is a noticeable improvement of 17.72% compared to the MFCC-based GMM algorithm in terms of error reduction.

• Speech Sciences
• /
• v.9 no.1
• /
• pp.39-47
• /
• 2002

This paper proposes an effective method to remedy the acoustic modeling problem inherent in the usual log-normal Parallel Model Composition intended for achieving robust speech recognition. In particular, the Gaussian kernels under the prescribed log-normal PMC cannot sufficiently express the corrupted speech distributions. The proposed scheme corrects this deficiency by judiciously selecting the 'fairly' corrupted component and by re-estimating it as a mixture of two distributions using data-driven PMC. As a result, some components become merged while equal number of components split. The determination for splitting or merging is achieved by means of measuring the similarity of the corrupted speech model to those of the clean model and the noise model. The experimental results indicate that the suggested algorithm is effective in representing the corrupted speech distributions and attains consistent improvement over various SNR and noise cases.