• Journal of the Korean Statistical Society
• /
• v.36 no.1
• /
• pp.1-29
• /
• 2007

Suppose one is trying to estimate a high dimensional vector of parameters from a series of one observation per parameter. Often, it is possible to take advantage of sparsity in the parameters by thresholding the data in an appropriate way. A marginal maximum likelihood approach, within a suitable Bayesian structure, has excellent properties. For very sparse signals, the procedure chooses a large threshold and takes advantage of the sparsity, while for signals where there are many non-zero values, the method does not perform excessive smoothing. The scope of the method is reviewed and demonstrated, and various theoretical, practical and computational issues are discussed, in particularly exploring the wide potential and applicability of the general approach, and the way it can be used within more complex thresholding problems such as curve estimation using wavelets.

• Journal of Digital Contents Society
• /
• v.19 no.1
• /
• pp.133-140
• /
• 2018

Accurate sentiment classification is an important research topic in sentiment analysis. This study suggests an efficient classification method of Korean sentiment using word2vec and ensemble methods which have been recently studied variously. For the 200,000 Korean movie review texts, we generate a POS-based BOW feature and a feature using word2vec, and integrated features of two feature representation. We used a single classifier of Logistic Regression, Decision Tree, Naive Bayes, and Support Vector Machine and an ensemble classifier of Adaptive Boost, Bagging, Gradient Boosting, and Random Forest for sentiment classification. As a result of this study, the integrated feature representation composed of BOW feature including adjective and adverb and word2vec feature showed the highest sentiment classification accuracy. Empirical results show that SVM, a single classifier, has the highest performance but ensemble classifiers show similar or slightly lower performance than the single classifier.

• The Korean Journal of Applied Statistics
• /
• v.11 no.1
• /
• pp.97-111
• /
• 1998

This paper suggests a Bayesian method for testing first-order markov correlation among linear regression disturbances. As a Bayesian test criterion, Bayes factor is derived in the form of generalized Savage-Dickey density ratio that is easily estimated by means of posterior simulation via Gibbs sampling scheme. Performance of the Bayesian test is evaluated and examined based upon a Monte Carlo experiment and an empirical data analysis. Efficiency of the posterior simulation is also examined.

• Journal of Korean Society of Transportation
• /
• v.29 no.1
• /
• pp.103-111
• /
• 2011

To evaluate highway safety countermeasures or identify high risk sites, the expected crashes for a site (or segment) have been estimated using the panel crash data. Past studies show that two different methods can be employed to estimate the expected crashes: observed crash based method and empirical Bayes (EB) method. This study conducts a simulation study to analyze how the estimation errors of the two estimates are affected by the different structures of the panel crash data and the presence of the change in safety over time. The results disclose that the estimation errors of the observed crash based estimates (i.e. the mean observed crash and comparative parallel estimate) are always greater than those of the EB estimates regardless of the structure of the panel crash data and the presence of the change in safety over time. Thus, it is highly recommended that the EB method be used in the study of traffic safety to obtain more reliable estimates for the expected crashes. In addition, this study corroborates that the estimation errors of the two estimates decrease as the analysis periods increase if safety does not change over time. Hence, it is also recommended that the 1-year analysis period used for identifying high risk sites in Korea be extended to produce more efficient estimates of the time-constant expected crashes.

• The Korean Journal of Applied Statistics
• /
• v.22 no.3
• /
• pp.617-626
• /
• 2009

A series of recent papers reported that the inter-gene correlations in Affymetrix microarray data sets were strong and long-ranged, and the assumption of independence or weak dependence among gene expression signals which was often employed without justification was in conflict with actual data. Qui et al. (2005) indicated that applying the nonparametric empirical Bayes method in which test statistics were pooled across genes for performing the statistical inference resulted in the large variance of the number of differentially expressed genes. Qui et al. (2005) attributed this effect to strong and long-ranged inter-gene correlations. Klebanov and Yakovlev (2007) demonstrated that the inter-gene correlations provided a rich source of information rather than being a nuisance in the statistical analysis and they developed, by transforming the original gene expression sequence, a sequence of independent random variables which they referred to as a ${\delta}$-sequence. We note in this report using two cDNA microarray data sets experimented in this country that the strong and long-ranged inter-gene correlations were still valid in cDNA microarray data and also the ${\delta}$-sequence of independence could be derived from the cDNA microarray data. This note suggests that the inter-gene correlations be considered in the future analysis of the cDNA microarray data sets.

• Health Policy and Management
• /
• v.22 no.2
• /
• pp.163-182
• /
• 2012

This study was performed to identify the difference of the area-based deprivation and the educational level on the cerebrovascular mortality in Korea. Data used in this study was obtained from the Death Certificate Data 2000 and the 2000 Census produced by Korean National Statistics(NSO). We classified the whole country into 246 areas based on the administrative districts. Then, the Standardized Mortality Ratio (SMR) in cerebrovascular disease was calculated according to the sex, education level and 246 areas. Its Predicted SMR was calculated by the Empirical Bayes Methods to reduce the variation of the SMR values. The area-based deprivation of 246 areas were measured using the modified Carstairs index in which the 5 indicators consisted of overcrowding, the unemployment ratio of men, the percentage of households classified low social class, the percentage of non home owners, and finally those houses lacking basic amenities. The correlation between the area-based deprivation and the SMR of the whole country and the correlation between the area-based deprivation and the SMR of each metropolitan cities or provinces was analyzed by the Pearson correlation analysis method. After classifying the deprivation of 246 areas into 5 levels, we performed the random intercept Poisson regression analysis after adjusting education level and age using Empirical Bayes Method to investigate the relationship between the 5 deprivation levels and the cerebrovascular mortality. The SMR was increased in lower education level. Each 246 areas had different values in SMR, Predicted SMR and area-based deprivation. The area-based deprivation and the SMR of the whole country was not correlated in both sexes. The education level of an individual was associated the risk of cerebrovascular mortality in men. The risk of cerebrovascular mortality increased with age compared to the reference(<30). The area-based deprivation was not associated with the risk of cerebrovascular in both sexes. The findings of this study suggest that the SMR had positive and negative correlations with area-based deprivation depending on the metropolitan cities or province. It also suggests that the individual education level and age were related with mortality and finally that the area-based deprivation was not associated to the cerebrovascular mortality in Korea.

• Journal of Korean Society of Transportation
• /
• v.27 no.3
• /
• pp.81-90
• /
• 2009

The safety management of a road network comprises four basic inter-related components：identification of sites(black spot) requiring safety investigation, diagnosis of safety problems, selection of feasible treatments for potential treatment candidates, and prioritization of treatments given limited budgets(Persaud, 2001). Identification process of selecting black spot is very important for efficient investigation of sites. In this study, the accident prediction model for EB method was developed by using accident data and geometric conditions of black spots selected from four-leg signalized intersections in In-cheon City for three years (2004-2006). In addition, by comparing the rank nomination technique using EB method to that by using accident counts, we managed to show the problems which the existing method have and the necessity for developing rational prediction model. As a result, in terms of total number of accidents, both the counts predicted by existing non-linear regression model and that by EB method have high good of fitness, but EB method, considering both the accident counts by sites and total number of accident, has better good of fitness than non-linear poison model. According to the result of the comparison of ranks nominated for treatment between two methods, the rank for treatment of almost sites does not change but SeoHae intersection and a few other intersections have significant changes in their rank. This shows that, with the technique proposed in the study, the RTM problem caused by using real accident counts can be overcome.

• Journal of Preventive Medicine and Public Health
• /
• v.54 no.5
• /
• pp.352-359
• /
• 2021

Objectives: The aim of this study was to examine the geographic distribution of diabetes mortality in Japan and identify socioeconomic factors affecting differences in municipality-specific diabetes mortality. Methods: Diabetes mortality data by year and municipality from 2013 to 2017 were extracted from Japanese Vital Statistics, and the socioeconomic characteristics of municipalities were obtained from government statistics. We calculated the standardized mortality ratio (SMR) of diabetes for each municipality using the empirical Bayes method and represented geographic differences in SMRs in a map of Japan. Multiple linear regression was conducted to identify the socioeconomic factors affecting differences in SMR. Statistically significant socioeconomic factors were further assessed by calculating the relative risk of mortality of quintiles of municipalities classified according to the degree of each socioeconomic factor using Poisson regression analysis. Results: The geographic distribution of diabetes mortality differed by gender. Of the municipality-specific socioeconomic factors, high rates of single-person households and unemployment and a high number of hospital beds were associated with a high SMR for men. High rates of fatherless households and blue-collar workers were associated with a high SMR for women, while high taxable income per-capita income and total population were associated with low SMR for women. Quintile analysis revealed a complex relationship between taxable income and mortality for women. The mortality risk of quintiles with the highest and lowest taxable per-capita income was significantly lower than that of the middle-income quintile. Conclusions: Socioeconomic factors of municipalities in Japan were found to affect geographic differences in diabetes mortality.

• Journal of the Korean Data and Information Science Society
• /
• v.24 no.4
• /
• pp.689-699
• /
• 2013

This paper studies the partial credibility application method by assuming the empirical prior or noninformative prior informations in auto insurnace business where intensive rating segmentation is expanded because of premium competition. Expanding of rating factor segmetation brings the increase of pricing cells, as a result, the number of cells for partial credibility application will increase correspondingly. This study is trying to suggest more accurate estimation method by considering the Bayesian framework. By using empirically well-known or noninformative information, inducing the proper posterior distribution and applying the Bayes estimate which is minimizing the error loss into the credibility method, we will show the advantage of Bayesian inference by comparison with current approaches. The comparison is implemented with square root rule which is a widely accepted method in insurance business. The convergence level towarding to the true risk will be compared among various approaches. This study introduces the alternative way of redcuing the error to the auto insurance business fields in need of various methods because of more segmentations.

• Bioinformatics and Biosystems
• /
• v.1 no.1
• /
• pp.67-72
• /
• 2006

The aim of this paper is to discuss the effect of missing values in detecting differentially expressed genes in a cDNA microarray experiment in the context of a one sample problem. We conducted a cDNA micro array experiment to detect differentially expressed genes for the metastasis of colorectal cancer based on twenty patients who underwent liver resection due to liver metastasis from colorectal cancer. Total RNAs from metastatic liver tumor and adjacent normal liver tissue from a single patient were labeled with cy5 and cy3, respectively, and competitively hybridized to a cDNA microarray with 7775 human genes. We used $M=log_2(R/G)$ for the signal evaluation, where Rand G denoted the fluorescent intensities of Cy5 and Cy3 dyes, respectively. The statistical problem comprises a one sample test of testing E(M)=0 for each gene and involves multiple tests. The twenty cDNA microarray data would comprise a matrix of dimension 7775 by 20, if there were no missing values. However, missing values occur for various reasons. For each gene, the no missing proportion (NMP) was defined to be the proportion of non-missing values out of twenty. In detecting differentially expressed (DE) genes, we used the genes whose NMP is greater than or equal to 0.4 and then sequentially increased NMP by 0.1 for investigating its effect on the detection of DE genes. For each fixed NMP, we imputed the missing values with K-nearest neighbor method (K=10) and applied the nonparametric t-test of Dudoit et al. (2002), SAM by Tusher et al. (2001) and empirical Bayes procedure by $L\ddot{o}nnstedt$ and Speed (2002) to find out the effect of missing values in the final outcome. These three procedures yielded substantially agreeable result in detecting DE genes. Of these three procedures we used SAM for exploring the acceptable NMP level. The result showed that the optimum no missing proportion (NMP) found in this data set turned out to be 80%. It is more desirable to find the optimum level of NMP for each data set by applying the method described in this note, when the plot of (NMP, Number of overlapping genes) shows a turning point.