• The Korean Journal of Applied Statistics
• /
• v.21 no.5
• /
• pp.825-833
• /
• 2008

Discriminant analysis based on Gaussian mixture models, an useful tool for multi-class classifications, can be extended to semi-supervised learning. We consider a model selection problem for a Gaussian mixture model in semi-supervised learning. More specifically, we adopt Bayesian information criterion to determine the number of subclasses in the mixture model. Through simulations, we illustrate the usefulness of the criterion.

• Proceedings of the Korean Institute of Electrical and Electronic Material Engineers Conference
• /
• 2010.06a
• /
• pp.49-49
• /
• 2010

In this paper, a full scaled gas discharge chamber was designed and fabricated for evaluating the dielectric performance of SF6/N2 mixture gases. And it describes work on AC and lightning impulse dielectric characteristics of SF6/N2 mixture insulation gas from experiments and full scale models.

• Proceedings of the IEEK Conference
• /
• 2002.06d
• /
• pp.195-198
• /
• 2002

We propose an efficient mixture Gaussian synthesis method for decision tree based state tying that produces better context-dependent models in a short period of training time. This method makes it possible to handle mixture Gaussian HMMs in decision tree based state tying algorithm, and provides higher recognition performance compared to the conventional HMM training procedure using decision tree based state tying on single Gaussian GMMs. This method also reduces the steps of HMM training procedure. We applied this method to training of PBS, and we expect to achieve a little point improvement in phoneme accuarcy and reduction in training time.

• Communications for Statistical Applications and Methods
• /
• v.25 no.6
• /
• pp.633-645
• /
• 2018

Gaussian error distributions are a common choice in traditional regression models for the maximum likelihood (ML) method. However, this distributional assumption is often suspicious especially when the error distribution is skewed or has heavy tails. In both cases, the ML method under normality could break down or lose efficiency. In this paper, we consider the log-concave and Gaussian scale mixture distributions for error distributions. For the log-concave errors, we propose to use a smoothed maximum likelihood estimator for stable and faster computation. Based on this, we perform comparative simulation studies to see the performance of coefficient estimates under normal, Gaussian scale mixture, and log-concave errors. In addition, we also consider real data analysis using Stack loss plant data and Korean labor and income panel data.

• 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.

• IEMEK Journal of Embedded Systems and Applications
• /
• v.10 no.2
• /
• pp.65-71
• /
• 2015

It has been known that GMM (Gaussian Mixture Model) based speaker identification systems using ML (Maximum Likelihood) and WMR (Weighting Model Rank) demonstrate very high performances. However, such systems are not so effective under practical environments, in terms of real time processing, because of their high calculation costs. In this paper, we propose a new speaker-pruning algorithm that effectively reduces the calculation cost. In this algorithm, we select 20% of speaker models having higher likelihood with a part of input speech and apply MWMR (Modified Weighted Model Rank) to these selected speaker models to find out identified speaker. To verify the effectiveness of the proposed algorithm, we performed speaker identification experiments using TIMIT database. The proposed method shows more than 60% improvement of reduced processing time than the conventional GMM based system with no pruning, while maintaining the recognition accuracy.

• Proceedings of the Membrane Society of Korea Conference
• /
• 1993.04a
• /
• pp.52-53
• /
• 1993

The pervaporation behavior of EtOH/Water mixture through IPN membranes was predicted in this study. The pervaporation characteristics of single polymer membrane were modeled according to the "six-coefficients model" proposed by Brun. In the case of the IPN membrane, two models were proposed according to the phase structure of the IPN. For a uniphase membrane with no phase separation, the compositional average of the single polymer membrane was used. in the case of the phase separated IPN's two cases existed. The first was the island and sea model: in which one phase was continuous and the other was dispersed. The second was the co-continuous model where two continuous phases existed. For these cases, the permeation rate and the separation factor of the IPN membrane were calculated using the experimental sorption data and the cornponent polymer properties. Comparison with the experimental data indicated that these models could be used to predict the performances of IPN membranes depending on the morphology of the IPN.

• Proceedings of the SAREK Conference
• /
• 2008.06a
• /
• pp.136-141
• /
• 2008

Due to environmental concerns $CO_2$ has been reintroduced as a potential candidate to replace HFCs in refrigeration systems since 1990s. In a refrigeration cycle, oil is utilized in lubricating a compressor. However, although oil separators are installed after a compressor oil is prone to leak to the whole system. The mixing of $CO_2$ and oil, even a small amount of oil, the heat transfer performance in heat exchanger deteriorated and the pressure drop inside tube increases. Therefore, it is needed to precisely estimate the mixture thermodynamic properties of $CO_2$-lubricant oil to correctly design a $CO_2$ refrigeration system. The commonly used method in estimating the mixture properties is the mole based weighting model. However, the accuracy of the method can not be assured. In the present study, $CO_2$-lubricant oil mixture properties including viscosity and density were estimated by using the mixture models, based on the equation of state (EOS).

• Journal of the Korean Data and Information Science Society
• /
• v.17 no.4
• /
• pp.1397-1403
• /
• 2006

This paper considers the testing problem for scale changes in autoregressive processes with heavy-tailed innovations. For a test, we propose the CUSUM test statistic based on the trimmed residuals. We perform a simulation study for the mixture normal and Cauchy innovations.

• Bulletin of the Korean Mathematical Society
• /
• v.43 no.4
• /
• pp.813-820
• /
• 2006

We consider a MAR model with ARCH type conditional heteroscedasticity. MAR-ARCH model can be derived as a smoothed version of the double threshold AR-ARCH model by adding a random error to the threshold parameters. Easy to check sufficient conditions for strict stationarity, ${\beta}-mixing$ property and existence of moments of the model are given via Markovian representation technique.