• The Korean Journal of Applied Statistics
• /
• v.24 no.6
• /
• pp.1213-1224
• /
• 2011

In high dimensionality where the number of variables are excessively larger than observations, it is required to remove the noninformative variables to cluster observations. Most model-based approaches for variable selection have been considered under the assumption of homoscedasticity and their models are mainly estimated by a penalized likelihood method. In this paper, a different approach is proposed to remove the noninformative variables effectively and to cluster based on the modified normal mixture model simultaneously. The validity of the model was provided and an EM algorithm was derived to estimate the parameters. Simulation studies and an experiment using real microarray dataset showed the effectiveness of the proposed method.

• Proceedings of the Korea Concrete Institute Conference
• /
• 2004.05a
• /
• pp.140-143
• /
• 2004

Heat of hydration of concrete under different curing temperatures can be characterized with knowledge of the thermal activity, the heat rate at the reference temperature, and the total heat of hydration of the mixture. The so-called multi-component hydration model incorporates the effect of following variables: cement chemical composition, cement fineness, secondary cementitious powders, mixture proportions, and concrete properties. However, the model does not consider the use of silica fume as a secondary cementitious powder. Therefore, the model that quantifies the heat of hydration due to the use of silica fume is needed. In this thesis, the effects of silica fume on heat of hydration are evaluated and the influence on the heat of hydration are also quantified to be included in the model, so that the analysis using modified multi-component hydration model for silica fume concrete provides more accurate results than normal concrete.

• Proceedings of the Korean Operations and Management Science Society Conference
• /
• 2007.11a
• /
• pp.469-473
• /
• 2007

For the quantitative analysis of inclusion using OES data, separating of noise and inclusion is needed. In previous methods assuming that noises come from a normal distribution, intensity levels beyond a specific threshold are determined as inclusions. However, it is not possible to classify inclusions in low intensity region using this method, even though every inclusion is an element of some chemical compound. In this paper, we assume that distribution of OES data is a Gaussian mixture and estimate the parameters of the mixture model using EM algorithm. Then, we calculate mixing ratio of noise and inclusion using these parameters to separate noise and inclusion.

• Preventive Nutrition and Food Science
• /
• v.14 no.4
• /
• pp.290-297
• /
• 2009

This study was carried out to investigate the dietary effects of chitosan mixture (CM), an herbal medicine mixture with Sukjihwang (HS), and CM containing HS (CHS) on obesity in an induced obese model of rats fed high-fat only (HF), in which supplemented diets of 5% CM (HCM), 5% HS (HHS), or 2.5% CM-2.5% HS (HCH) was tested for 6 weeks. Body weight gains, obesity indexes, and body fat contents in the experimental groups (HCM, HHS, HCH) were decreased compared with HF group. The levels of serum triglyceride, total lipid, total cholesterol and LDL-cholesterol in the experimental groups were markedly decreased, however HDL-cholesterol levels in the experimental groups were slightly increased compared with HF group. In addition, although serum ALT and AST activity, and relative organ weights were lower than those of HF group, serum albumin contents were not significantly different in all experimental groups including the normal control group (NC). In conclusion, there are improved effects on obesity in the obese model of animals with all experimental diets supplementations, and the improvement degrees on obesity depend on the content and compositions of the herbal medicine mixture. Further study is needed on the anti-obesity mechanism of these diets.

• MALSORI
• /
• no.64
• /
• pp.77-88
• /
• 2007

Nowadays, mel-frequency cesptral coefficients (MFCCs) and Gaussian mixture models (GMMs) are used for the pathological voice detection. This paper suggests a method to improve the performance of the pathological/normal voice classification based on the MFCC-based GMM. We analyze the characteristics of the mel frequency-based filterbank energies using the fisher discriminant ratio (FDR). And the feature vectors through the linear discriminant analysis (LDA) transformation of the filterbank energies (FBE) and the MFCCs are implemented. An accuracy is measured by the GMM classifier. This paper shows that the FBE LDA-based GMM is a sufficiently distinct method for the pathological/normal voice classification, with a 96.6% classification performance rate. The proposed method shows better performance than the MFCC-based GMM with noticeable improvement of 54.05% in terms of error reduction.

• Communications for Statistical Applications and Methods
• /
• v.16 no.4
• /
• pp.573-578
• /
• 2009

We prove the consistency of a penalized maximum likelihood estimate proposed by Ahn (2001). The PMLE not only avoids the well-known problem that the ordinary likelihood of the normal mixture model is unbounded for any given sample size, but also removes redundant components.

• Journal of Periodontal and Implant Science
• /
• v.28 no.2
• /
• pp.377-388
• /
• 1998

This double-blind controlled clinical and microbiological study was carried out to determine the effects of mouthwash preparation containing the mixture of herbal extract on developing plaque and gingivitis in the experimental gingivitis model. Following a 2-week normalization period, 34 dental students were distributed randomly into 1 of 3 treatment groups. They rinsed, under supervision, two times daily for 3 weeks with either normal saline(CT), 0.1% chlorhexidine(CH), or the mixture of herbal extract (HT), but refrained from any oral hygiene measures. The Plaque Index(PlI), the Gingival Index(GI), and the amount of Gingival Crevicular Fluid(GCF) were measured at week 0,1,2, and 3 of the experimental period, while the assessment of total wet weight of plaque and the phase contrast microscopic examination of plaque were performed at the end of experimental period(3 weeks). Subjects using mouthrinse preparation containing the mixture of herbal extract demonstrated negligible, if any, changes in the accumulation and microbial composition of plaque compared to those using normal saline, while the reduction of gingival inflammation by this mixture was highly significant and comparable to that of chlorhexidine. The results of this study indicate that the preparation containing the mixture of herbal extract do not provide any antiplaque benefits but is very effective in inhibiting the development of and in reducing existing experimental gingivitis when used as mouthrinse. Further research is needed to determine whether a significant reduction of gingival inflammation without a concomitant decrease in plaque accumulation is of clinical importance.

• The Korean Journal of Applied Statistics
• /
• v.25 no.6
• /
• pp.1037-1047
• /
• 2012

The generalized linear mixed model(GLMM) is widely used in fitting categorical responses of clustered data. In the numerical approximation of likelihood function the normality is assumed for the random effects distribution; subsequently, the commercial statistical packages also routinely fit GLMM under this normality assumption. We may also encounter departures from the distributional assumption on the response variable. It would be interesting to investigate the impact on the estimates of parameters under misspecification of distributions; however, there has been limited researche on these topics. We study the sensitivity or robustness of the maximum likelihood estimators(MLEs) of GLMM for counts data when the true underlying distribution is normal, gamma, exponential, and a mixture of two normal distributions. We also consider the effects on the MLEs when we fit Poisson-normal GLMM whereas the outcomes are generated from the negative binomial distribution with overdispersion. Through a small scale Monte Carlo study we check the empirical coverage probabilities of parameters and biases of MLEs of GLMM.

• 한국연소학회:학술대회논문집
• /
• 2002.06a
• /
• pp.123-132
• /
• 2002

A numerical study was conducted to investigate the combustion phenomena of normal start and unstart processes based on ISL's RAMAC 30 experiments with different diluent amounts and fill pressures in a ram accelerator. The initial projectile launching speed was 1.8 km/s which corresponded to the superdetonative speed of the stoichiometric $H_2/O_2$ mixture diluted with 5 $CO_2$ or 4 $CO_2$. Experiments with same condition except for projectile surface material demonstrated that ignition was successful with an aluminum projectile, but no combustion was observed in case of a steel projectile. In this study, it was found that neither shock nor viscous heating was sufficient to ignite the mixture at a low speed of 1.8 km/s, as was found in the experiments using a steel projectile. However, we could succeed in igniting the mixtures by imposing a minimal amount of additional heat to the combustor section and simulate the normal start and unstart processes found in the experiments with an aluminum projectile. For the numerical simulation of supersonic combustion, multi-species Navier-Stokes equations coupled with a Baldwin-Lomax turbulence model and detailed chemistry reaction equations of $H_2/O_2/CO_2$ suitable for high-pressure gaseous combustion were considered. The governing equations were discretized by a high order accurate upwind scheme and solved in a fully coupled manner with a fully implicit, time accurate integration method. The numerical results matched almost exactly to the experimental results. As a result, it was found that the normal start and unstart processes depended on the strength of gas mixture, development of shock-induced combustion wave stabilized by the first separation bubble, and its size and location.

• Communications for Statistical Applications and Methods
• /
• v.18 no.3
• /
• pp.391-401
• /
• 2011

Consider a cluster, so called a 'son cluster', whose mean and variance is composed of the means and variances of both clusters called as a 'father cluster' and a 'mother cluster'. In this paper, a method for identifying each of three clusters is provided by modeling the relationship with father and mother clusters. Under the normal mixture model, the parameters are estimated via EM algorithm. We were able to overcome the problems of estimation using ECM approximation. Numerical examples show that our method can effectively identify the three clusters, so called a 'family of clusters'.