Go to the main menu
Skip to content
Go to bottom
REFERENCE LINKING PLATFORM OF KOREA S&T JOURNALS
> Journal Vol & Issue
Journal of the Korean Data and Information Science Society
Journal Basic Information
Journal DOI :
Korean Data and Information Science Society
Editor in Chief :
Volume & Issues
Volume 27, Issue 4 - Jul 2016
Volume 27, Issue 3 - May 2016
Volume 27, Issue 2 - Mar 2016
Volume 27, Issue 1 - Jan 2016
Selecting the target year
Electricity forecasting model using specific time zone
Shin, YiRe ; Yoon, Sanghoo ;
Journal of the Korean Data and Information Science Society, volume 27, issue 2, 2016, Pages 275~284
DOI : 10.7465/jkdi.2016.27.2.275
Accurate electricity demand forecasts is essential in reducing energy spend and preventing imbalance of the power supply. In forcasting electricity demand, we considered double seasonal Holt-Winters model and TBATS model with sliding window. We selected a specific time zone as the reference line of daily electric demand because it is least likely to be influenced by external factors. The forecasting performance have been evaluated in terms of RMSE and MAPE criteria. We used the observations ranging January 4, 2009 to December 31 for testing data. For validation data, the records has been used between January 1, 2012 and December 29, 2012.
Study on the determinants of employment duration in the youth-intern project
Park, Sungik ; Ryu, Jangsoo ; Kim, Jonghan ; Cho, Jangsik ;
Journal of the Korean Data and Information Science Society, volume 27, issue 2, 2016, Pages 285~294
DOI : 10.7465/jkdi.2016.27.2.285
In general, employment duration is influenced by the individual characteristics (level-1) as well as type of the occupational characteristics (level-2). That is, the data has hierarchical structure in the sense that individual employment duration is influenced by the individual-level variables (level-1) and the job-level (level-2) variables. In this paper, we study the determinants of the employment duration of youth-intern in the SMEs (small and medium enterprises) using Cox's mixed effect model. Major results at level-1 variables are as followings. First, the hazard rate of treatment group is lower than that of control group. Second, the hazard rate of woman is lower than that of man. Also, the hazard rate is lower, for the older and the workers working in the bigger company. Investigation of level-2 variables has shown that random effect for job-level is statistically significant.
A study on the factor in a view of mathematical learning
Kim, Sang-Lyong ;
Journal of the Korean Data and Information Science Society, volume 27, issue 2, 2016, Pages 295~304
DOI : 10.7465/jkdi.2016.27.2.295
This study investigates significant factors in mathematical learning and examines the inter-grade and gender-based differences of elementary students. Five factors that are counted to affect the view of mathematical learning are (1) confidence, (2) utility, (3) aversion, (4) practical ability, and (5) traditional view of mathematical learning. The factor analyses on third graders and sixth graders each illustrate the features of inter-grade factors. The result also indicates that the factors may vary depending on the traits and circumstances of students surveyed. Third graders are more likely to be positive compared to sixth graders in terms of confidence and practical ability, which calls for implementing 'doing mathematics' and reinforcing the method of mathematical learning in the general educational field.
A study on fractal dimensions of art works
Synn, Chaeki F. ; Heo, A-Young ; Kim, Seul Gee ; Park, Cheolyong ;
Journal of the Korean Data and Information Science Society, volume 27, issue 2, 2016, Pages 305~314
DOI : 10.7465/jkdi.2016.27.2.305
In this study, an analysis is performed for comparing the fractal dimension of Jackson Pollock's art works with that of Korean Infomel art works. In order to test the hypothesis that Jackson Pollock's fractal dimension is different from Korean Informel's, data is collected for the fractal dimensions of 30 Jackson Pollock's and 45 Korean Informel art works. The results show that Korean Informel's fractal dimension is larger than Jackson Pollock's. This might be interpreted that the pattern (in finer scale) of Korean Informel art works is closer to planes, rather than lines or points, compared to that of Jackson Pollock's.
Bias caused by nonresponses and suggestion for increasing response rate in the telephone survey on election
Heo, Sunyeong ; Yi, Sucheol ;
Journal of the Korean Data and Information Science Society, volume 27, issue 2, 2016, Pages 315~325
DOI : 10.7465/jkdi.2016.27.2.315
Thanks to the advantages of low cost and quick results, public opinion polls on election in Korea have been generally conducted by telephone survey, even though it has critical disadvantage of low response rate. In public opinion polls on election in Korea, the general method to handle nonresponses is adjusting the survey weight to estimate parameters. This study first drives mathematical expression of estimator and its bias with variance estimators with/without nonresponses in election polls in Korea. We also investigates the nonresponse rate of telephone survey on 2012 Korea presidential election. The average response rate was barely about 14.4%. In addition, we conducted a survey in April 2014 on the respondents's attitude toward telephone surveys. In the survey, the first reason for which respondents do not answer on public opinion polls on election was "feel bothered". And the aged 20s group, the most low response group, also gave the same answer. We here suggest that survey researchers motivate survey respondents, specially younger group, to participate surveys and find methods boosting response rate such as giving incentive.
A deep learning analysis of the Chinese Yuan's volatility in the onshore and offshore markets
Lee, Woosik ; Chun, Heuiju ;
Journal of the Korean Data and Information Science Society, volume 27, issue 2, 2016, Pages 327~335
DOI : 10.7465/jkdi.2016.27.2.327
The People's Republic of China has vigorously been pursuing the internationalization of the Chinese Yuan or Renminbi after the financial crisis of 2008. In this view, an abrupt increase of use of the Chinese Yuan in the onshore and offshore markets are important milestones to be one of important currencies. One of the most frequently used methods to forecast volatility is GARCH model. Since a prediction error of the GARCH model has been reported quite high, a lot of efforts have been made to improve forecasting capability of the GARCH model. In this paper, we have proposed MLP-GARCH and a DL-GARCH by employing Artificial Neural Network to the GARCH. In an application to forecasting Chinese Yuan volatility, we have successfully shown their overall outperformance in forecasting over the GARCH.
Proposition of polytomous discrimination index and test statistics
Choi, Jin Soo ; Hong, Chong Sun ;
Journal of the Korean Data and Information Science Society, volume 27, issue 2, 2016, Pages 337~351
DOI : 10.7465/jkdi.2016.27.2.337
There exist many real situations that statistical decision problems are classified into more than two categories. In these cases, the concordance statistics by the pair approach are mostly used. However, the expression of the classification of categories are ambiguous. Recently, the standardized evaluation data and re-expressed concordance statistics are defined and could be explained their meanings. They have still some non-specific problems for standard criteria of the statistics. Since these can be considered between result and truth categories additionally, two alternative concordance statistics might be proposed in this paper. Some advantages are founded that the proposed statistics could be discriminated all possible cases for two randomly selected categories. Moreover since the proposed statistics are represented with indicator functions, these could be transformed non-parametrically, so that these concordances are used for hypothesis testing.
Signed Hellinger measure for directional association
Park, Hee Chang ;
Journal of the Korean Data and Information Science Society, volume 27, issue 2, 2016, Pages 353~362
DOI : 10.7465/jkdi.2016.27.2.353
By Wikipedia, data mining is the process of discovering patterns in a big data set involving methods at the intersection of association rule, decision tree, clustering, artificial intelligence, machine learning. and database systems. Association rule is a method for discovering interesting relations between items in large transactions by interestingness measures. Association rule interestingness measures play a major role within a knowledge discovery process in databases, and have been developed by many researchers. Among them, the Hellinger measure is a good association threshold considering the information content and the generality of a rule. But it has the drawback that it can not determine the direction of the association. In this paper we proposed a signed Hellinger measure to be able to interpret operationally, and we checked three conditions of association threshold. Furthermore, we investigated some aspects through a few examples. The results showed that the signed Hellinger measure was better than the Hellinger measure because the signed one was able to estimate the right direction of association.
Choosing clusters for two-stage household surveys
Park, Inho ;
Journal of the Korean Data and Information Science Society, volume 27, issue 2, 2016, Pages 363~372
DOI : 10.7465/jkdi.2016.27.2.363
Two-stage sample designs are commonly used for household surveys in Korea using as clusters the enumeration districts (EDs). Since clustering decomposes the population variation into within- and between-cluster variations, the sample sizes allocated in stages can affect the overall precision. Alternative clusters are often considered due to diverse reasons such as the EDs' limitation in size, being out-of-date, and in-assessibility to their household lists. In addition, the EDs are currently under development by the Statistics Korea as an joint effort toward their transition from the traditional practice to the register census from 2015. We present an approach for evaluating the difference in the precision of the mean estimators of the sets of the cluster units in between a hierachical and nested form, where the design effect is used to reflect the effect of the clustering and the sample allocation. We also demonstrate our approach using the U.S. Census counts from the year 2000 for Anne Arundel County in Maryland. Our research shows that the within-cluster variance can be significantly different for survey variables and thus the choice of cluster units and the associated sample allocation scheme should reflect the corresponding variance decomposition due to clustering.
A simulation comparison on the analysing methods of Likert type data
Kim, Hyun Chul ; Choi, Seung Kyoung ; Choi, Dong Ho ;
Journal of the Korean Data and Information Science Society, volume 27, issue 2, 2016, Pages 373~380
DOI : 10.7465/jkdi.2016.27.2.373
Even though Likert type data is ordinal scale, many researchers who regard Likert type data as interval scale adapt as parametric methods. In this research, simulations have been used to find out a proper analysis of Likert type data. The locations and response distributions of five point Likert type data samples having diverse distribution have been evaluated. In estimating samples' locations, we considered parametric method and non-parametric method, which are t-test and Mann-Whitney test respectively. In addition, to test response distribution, we employed Chi-squared test and Kolmogorov-Smirnov test. In this study, we assessed the performance of the four aforementioned methods by comparing Type I error ratio and statistical power.
BCDR algorithm for network estimation based on pseudo-likelihood with parallelization using GPU
Kim, Byungsoo ; Yu, Donghyeon ;
Journal of the Korean Data and Information Science Society, volume 27, issue 2, 2016, Pages 381~394
DOI : 10.7465/jkdi.2016.27.2.381
Graphical model represents conditional dependencies between variables as a graph with nodes and edges. It is widely used in various fields including physics, economics, and biology to describe complex association. Conditional dependencies can be estimated from a inverse covariance matrix, where zero off-diagonal elements denote conditional independence of corresponding variables. This paper proposes a efficient BCDR (block coordinate descent with random permutation) algorithm using graphics processing units and random permutation for the CONCORD (convex correlation selection method) based on the BCD (block coordinate descent) algorithm, which estimates a inverse covariance matrix based on pseudo-likelihood. We conduct numerical studies for two network structures to demonstrate the efficiency of the proposed algorithm for the CONCORD in terms of computation times.
Analysis of error source in subjective evaluation results on Taekwondo Poomsae: Application of generalizability theory
Cho, Eun Hyung ;
Journal of the Korean Data and Information Science Society, volume 27, issue 2, 2016, Pages 395~407
DOI : 10.7465/jkdi.2016.27.2.395
This study aims to apply the G-theory for estimation of reliability of evaluation scores between raters on Taekwondo Poomsae rating categories. Selecting a number of game days and raters as multiple error sources, we analyzed the error sources caused by relative magnitude of error variances of interaction between the factors and proceeded with D-study based on the results of G-study for optimal determination of measurement condition. The results showed below. The estimated outcomes of variance component for accuracy among the Taekwondo Poomsae categories with G-theory showed that impact of error was the biggest influence factor in raters conditions and in order of interaction in subjects and between subjects, also impact of variance component estimation error on expression category was the major influence factor in interaction and in order of the between subjects and raters. Finally, the result of generalizability coefficient estimation via D-study showed that measurement condition of optimal level depend on the number of raters was 8 persons of raters on accuracy category, and stable reliability on expression category was gained when the raters were 7 persons.
The effects on fatigue and accuracy of cardiopulmonary resuscitation of the verbal-order method based on different time intervals (3, 4 minutes)
Lee, Mi Kyoung ; Yang, Jeong Ok ; Jung, Joo Ha ; Lee, Kyeong Jun ; Cho, Youngseuk ;
Journal of the Korean Data and Information Science Society, volume 27, issue 2, 2016, Pages 409~417
DOI : 10.7465/jkdi.2016.27.2.409
The purpose of this study was to demonstrate the effect on the degree of fatigue and accuracy of cardiopulmonary resuscitation according to the different time delays (3 minutes, 4 minutes). Carrying out repeated measures of variance (repeated ANOVA), we have shown that time effect (F = 7.835, p <.01) and group effect (F = 8.695, p<.01) and the interaction effect between time and group (F = 12.582, p<.001) were all statistically significant. It means, in the test of the main effect of group and time (3 minutes, 4 minutes) using the Bonferroni method, it turned out that the amount of lactic acid of the experimental group was larger than that of the control group (p<.01), and there was no difference until 3 minutes, but the difference of the amount of lactic acid was shown between before the experiment and after 4 minutes, and between 3 minutes and 4 minutes (p <.05), respectively. Then, in the result of the corresponding sample t-test, for comparing the according to the measurement time, the accuracy after 3 minutes became higher than the case of 4 minutes (t = 4.584, p <.001). Therefore, before 119 arrives performing cardiopulmonary resuscitation for emergency, rescuers need to perform cardiopulmonary resuscitation alternating with others before 3 minutes.
Effect on ambulatory dental visitation frequency according to pack-years of smoking
Jeong, Sun-Rak ; Doo, Young-Taek ; Lee, Won Kee ;
Journal of the Korean Data and Information Science Society, volume 27, issue 2, 2016, Pages 419~427
DOI : 10.7465/jkdi.2016.27.2.419
To examine whether the effect on utilization of ambulatory dental care are associated with oral disease according to pack-years of smoking in Korean population. Using data from Korea Health Panel between 2008 and 2012, we analyzed 3,866 participants who were male and more than 20 years. Pack-years of smoking were significantly associated with utilization in ambulatory dental care after adjustment for age, marital status, family income, and chronic disease. Ambulatory dental visitation frequency has been estimated to increase by 6% when 10.0 pack-years of smoking increased. Especially, the smokers who had 20.0~29.9 and 30.0 or more pack-years of smoking in forties and fifties males were 25% and 52% respectively more than non-smokers in utilization of ambulatory dental care.
Changes rate in selection of Yorkshire pig for productive traits using the integrated test records among GGPs
Cho, Kwang-Hyun ; Kim, Sung-Hoon ; Park, Kyung-Do ;
Journal of the Korean Data and Information Science Society, volume 27, issue 2, 2016, Pages 429~435
DOI : 10.7465/jkdi.2016.27.2.429
Heritability estimates for daily gain (g), backfat thickness (mm), days to 90kg (day), loin eye depth (mm) and meat percent (%) were 0.40, 0.44, 0.40, 0.25 and 0.48, respectively. Estimates of correlation between breeding value and rank for meat productivity traits by model 1 and 2 were 0.995 1.000 and 0.991 1.000, respectively and highly significant (p< 0.0001), and they were almost identical to the breeding values estimated by different farms. When top 5% and top 10% animals were selected by meat productive traits at different farms, markedly different animals were selected by farms since the selected improvement traits in each farm maintaining closed herds were different. Therefore, it seems to be desirable that superior pigs should be selected after the establishment of evaluation system for genetic performance at national level using the integrated data obtained from various farms.
An analysis of determinants of individual's pension purchase using structural equation model
Lee, Chanhee ; Chun, Heuiju ;
Journal of the Korean Data and Information Science Society, volume 27, issue 2, 2016, Pages 437~449
DOI : 10.7465/jkdi.2016.27.2.437
This study empirically analyzes casual relationships among psychological and financial factors influencing the subscription of individual pension and identifies mediation effect by the structural equation model. The analysis based on survey data (N=928) from common people living in the seven major cities in Korea shows that potential longevity, coping efficacy to the financial risks from longevity, subjective knowledge on longevity (cognitive factor), trust in insurers (social factor), and financial literacy(financial factor) are key players exercising direct influence over the subscription of individual pension. It also has been shown that affect on longevity, potential longevity, trust in government, and coping efficacy influence the emotions on longevity, fear and worry, and emotions on longevity are not significant in explaining the subscription of individual pension. Trust in insurers and financial knowledge of prospects have also been reaffirmed to be crucial factors in the activation of individual pension subscription.
The relationship among critical thinking disposition, nursing process competency and evidence-based practice competency in nurses working in hospitals
Kim, Kyoung Yun ; Lee, Eunjoo ;
Journal of the Korean Data and Information Science Society, volume 27, issue 2, 2016, Pages 451~461
DOI : 10.7465/jkdi.2016.27.2.451
The purpose of this study was to identify relationship among critical thinking, nursing process competency and evidence based practice competency of nurses working in middle sized hospitals. A descriptive correlational study design was used. The data were collected from 262 nurses working in the three hospitals located in G and P city using self-administered questionnaires. Data were analyzed using descriptive statistic, independent t-test, one-way ANOVA with scheffe, Pearson's correlation coefficient, and hierarchical multiple linear regression using SPSS Statistics 21.0 program. Nurses' evidence-based practice competency had positive correlation with critical thinking disposition (r=.42, p <.001) and nursing process competency (r=.58, p <.001). Full model accounted for 65.1% of evidence-based practice competency by nursing process competency (
, p <.001), critical thinking disposition (
, p <.001), critical thinking educational experience (
, p=.004), Master's degree (
, p=.032). Conclusion: o increase evidence-based practice competency of nurses, it is required to develop critical thinking disposition and nursing process competency of nurses.
The relationship among nursing student's knowledge, nursing skill and perceived performance of tracheostomy care
Lee, Sun Hee ; Kim, Soon Hee ;
Journal of the Korean Data and Information Science Society, volume 27, issue 2, 2016, Pages 463~475
DOI : 10.7465/jkdi.2016.27.2.463
This study was to search the relationship among knowledge, nursing skill, perceived performance in graduating nursing class. The participants were 90 members of to graduating nursing class in D city. Data were collected from September 4, 2014 to September 22, 2014 using a questionnaire and core nursing skills checklists. Data analysis was done with SPSS/WIN 23.0 using descriptive statistics and Pearson correlation. It found that knowledge of hand washing had a positive correlation between the nursing skill and perceived performance. The most vulnerablenursing skill was hand washing. The best nursing skill was sterilization and withdrawing contaminated products from the patient (Place the inner tube immersed in a solution of hydrogen peroxide). Based on the findings of this study, activating prior knowledge needs to be stressed. Thus, it would be necessary to include more effective motivation in designing experiential education program for cognitive performance.
Testing of risk perception survey - Diabetes mellitus in Korea
Kang, Soo Jin ;
Journal of the Korean Data and Information Science Society, volume 27, issue 2, 2016, Pages 477~486
DOI : 10.7465/jkdi.2016.27.2.477
This study was to evaluate of the reliability and validity of the Risk Perception Survey - Diabetes Mellitus (RPS-DM) with Korean diabetes patients. A total of 183 patients participated in this study from December 4, 2014 to January 1, 2015 with self-reported questionnaires. The data was analyzed using exploratory factor analysis, cronbach's alpha, and item to total correlation. The factor structure of the instrument showed the cumulative variance of 45.1% in the factor analysis and a four-factor structure was found to be appropriate. The comparative site risk score matched with the RPS-DM in English except item 7, 8, and 12. The RPS-DM in Korean version has been found to be reliable and valid.
The effects of education for hemodialysis patients with a family caregiver on self-care practice and blood biochemical parameters
Park, Ji Hyun ; Choi, Hyunkyung ;
Journal of the Korean Data and Information Science Society, volume 27, issue 2, 2016, Pages 487~498
DOI : 10.7465/jkdi.2016.27.2.487
The purpose of this study was to identify the effects of education for hemodialysis patient with a family caregiver on self-care practice and on blood biochemical parameters. A nonequivalent control group pretest and posttest design has been employed for analysis. Among hemodialysis patients in C university hospital located in Daegu, 56 subjects, 28 in experimental and 28 in control group, participated in the study from April to May in 2015. The experimental group whose family caregiver participated in education for hemodialysis patient reported significant differences in self-care practice (t=3.36, p=.001) and serum potassium level (t=2.78, p=.007) from the control group whose family caregiver did not participate. Considering positive effects of family education, we suggest education programs incorporating family caregivers be used as an effective nursing intervention.
Analysis of multi-center bladder cancer survival data using variable-selection method of multi-level frailty models
Kim, Bohyeon ; Ha, Il Do ; Lee, Donghwan ;
Journal of the Korean Data and Information Science Society, volume 27, issue 2, 2016, Pages 499~510
DOI : 10.7465/jkdi.2016.27.2.499
It is very important to select relevant variables in regression models for survival analysis. In this paper, we introduce a penalized variable-selection procedure in multi-level frailty models based on the "frailtyHL" R package (Ha et al., 2012). Here, the estimation procedure of models is based on the penalized hierarchical likelihood, and three penalty functions (LASSO, SCAD and HL) are considered. The proposed methods are illustrated with multi-country/multi-center bladder cancer survival data from the EORTC in Belgium. We compare the results of three variable-selection methods and discuss their advantages and disadvantages. In particular, the results of data analysis showed that the SCAD and HL methods select well important variables than in the LASSO method.
Predictors of subjectives happiness for male nursing students
Park, Ji-Hyun ; Jo, Geum-Yi ;
Journal of the Korean Data and Information Science Society, volume 27, issue 2, 2016, Pages 511~522
DOI : 10.7465/jkdi.2016.27.2.511
The aim of this study was to identify the significant predictors of subjective happiness for male nursing students. Collecting data from 171 male nursing students, we carried out t-test, ANOVA, Scheffe test, Pearson correlation coefficient, and stepwise multiple regression on SPSS Win 20.0. The significant predictors of subjective happiness for male nursing students were self-esteem, gratitude disposition, perceived relationship of peer, and flow. These factors explained 56.1% of subjective happiness. The results suggest that an effective approach to happiness promotion program for male nursing students should consider self-esteem, gratitude disposition, perceived relationship of peer, and flow.
Multioutput LS-SVR based residual MCUSUM control chart for autocorrelated process
Hwang, Changha ;
Journal of the Korean Data and Information Science Society, volume 27, issue 2, 2016, Pages 523~530
DOI : 10.7465/jkdi.2016.27.2.523
Most classical control charts assume that processes are serially independent, and autocorrelation among variables makes them unreliable. To address this issue, a variety of statistical approaches has been employed to estimate the serial structure of the process. In this paper, we propose a multioutput least squares support vector regression and apply it to construct a residual multivariate cumulative sum control chart for detecting changes in the process mean vector. Numerical studies demonstrate that the proposed multioutput least squares support vector regression based control chart provides more satisfying results in detecting small shifts in the process mean vector.
Geographically weighted kernel logistic regression for small area proportion estimation
Shim, Jooyong ; Hwang, Changha ;
Journal of the Korean Data and Information Science Society, volume 27, issue 2, 2016, Pages 531~538
DOI : 10.7465/jkdi.2016.27.2.531
In this paper we deal with the small area estimation for the case that the response variables take binary values. The mixed effects models have been extensively studied for the small area estimation, which treats the spatial effects as random effects. However, when the spatial information of each area is given specifically as coordinates it is popular to use the geographically weighted logistic regression to incorporate the spatial information by assuming that the regression parameters vary spatially across areas. In this paper, relaxing the linearity assumption and propose a geographically weighted kernel logistic regression for estimating small area proportions by using basic principle of kernel machine. Numerical studies have been carried out to compare the performance of proposed method with other methods in estimating small area proportion.
Multivariate CUSUM control charts for monitoring the covariance matrix
Choi, Hwa Young ; Cho, Gyo-Young ;
Journal of the Korean Data and Information Science Society, volume 27, issue 2, 2016, Pages 539~548
DOI : 10.7465/jkdi.2016.27.2.539
This paper is a study on the multivariate CUSUM control charts using three different control statistics for monitoring covariance matrix. We get control limits and ARLs of the proposed multivariate CUSUM control charts using three different control statistics by using computer simulations. The performances of these proposed multivariate CUSUM control charts have been investigated by comparing ARLs. The purpose of control charts is to detect assignable causes of variation so that these causes can be found and eliminated from process, variability will be reduced and the process will be improved. We show that the charts based on three different control statistics are very effective in detecting shifts, especially shifts in covariances when the variables are highly correlated. When variables are highly correlated, our overall recommendation is to use the multivariate CUSUM control charts using trace for detecting changes in covariance matrix.
LS-SVM for large data sets
Park, Hongrak ; Hwang, Hyungtae ; Kim, Byungju ;
Journal of the Korean Data and Information Science Society, volume 27, issue 2, 2016, Pages 549~557
DOI : 10.7465/jkdi.2016.27.2.549
In this paper we propose multiclassification method for large data sets by ensembling least squares support vector machines (LS-SVM) with principal components instead of raw input vector. We use the revised one-vs-all method for multiclassification, which is one of voting scheme based on combining several binary classifications. The revised one-vs-all method is performed by using the hat matrix of LS-SVM ensemble, which is obtained by ensembling LS-SVMs trained using each random sample from the whole large training data. The leave-one-out cross validation (CV) function is used for the optimal values of hyper-parameters which affect the performance of multiclass LS-SVM ensemble. We present the generalized cross validation function to reduce computational burden of leave-one-out CV functions. Experimental results from real data sets are then obtained to illustrate the performance of the proposed multiclass LS-SVM ensemble.
Remarks on correlated error tests
Kim, Tae Yoon ; Ha, Jeongcheol ;
Journal of the Korean Data and Information Science Society, volume 27, issue 2, 2016, Pages 559~564
DOI : 10.7465/jkdi.2016.27.2.559
The Durbin-Watson (DW) test in regression model and the Ljung-Box (LB) test in ARMA (autoregressive moving average) model are typical examples of correlated error tests. The DW test is used for detecting autocorrelation of errors using the residuals from a regression analysis. The LB test is used for specifying the correct ARMA model using the first some sample autocorrelations based on the residuals of a tted ARMA model. In this article, simulations with four data generating processes have been carried out to evaluate their performances as correlated error tests. Our simulations show that the DW test is severely dependent on the assumed AR(1) model but isn't sensitive enough to reject the misspecified model and that the LB test reports lackluster performance in general.
Noninformative priors for linear combinations of exponential means
Lee, Woo Dong ; Kim, Dal Ho ; Kang, Sang Gil ;
Journal of the Korean Data and Information Science Society, volume 27, issue 2, 2016, Pages 565~575
DOI : 10.7465/jkdi.2016.27.2.565
In this paper, we develop the noninformative priors for the linear combinations of means in the exponential distributions. We develop the matching priors and the reference priors. The matching priors, the reference prior and Jeffreys' prior for the linear combinations of means are developed. It turns out that the reference prior and Jeffreys' prior are not a matching prior. We show that the proposed matching prior matches the target coverage probabilities much more accurately than the reference prior and Jeffreys' prior in a frequentist sense through simulation study, and an example based on real data is given.