비대칭 라플라스 분포를 이용한 분위수 회귀

박혜정,Park, Hye-Jung 한국데이터정보과학회 2009 한국데이터정보과학회지 Vol.20 No.6

분위수 회귀모형은 확률변수들 사이에 확률적인 관계구조를 포함한 함수 모형을 좀 더 완벽하게 추정하도록 제공한다. 본 논문에서는 함수 추정에 로버스트하다고 알려져 있는 서포트벡터기계 기법과 이중벌칙커널기계를 이용하여 분위수 회귀모형을 추정하고자 한다. 이중벌칙커널기계는 고차원의 입력변수에 대한 분위수 회귀가 요구될 때 분위수 회귀모형을 잘 추정한다고 알려져 있다. 또한 본 논문에서는 광범위한 형태의 분위수 회귀모형 추정을 위해서 정규분포보다 비대칭 라플라스 분포를 이용한다. 본 논문에서 제안한 모형은 분위수 회귀모형 추정을 위해서 서포트벡터기계 기법에 이중벌칙커널기계를 이용하여 각각의 평균과 분산을 동시에 추정한다. 평균과 분산함수 추정을 위해 사용된 커널함수의 모수들은 최적의 값을 찾기 위해 일반화근사 교차타당성을 이용한다. Quantile regression has become a more widely used technique to describe the distribution of a response variable given a set of explanatory variables. This paper proposes a novel modelfor quantile regression using doubly penalized kernel machine with support vector machine iteratively reweighted least squares (SVM-IRWLS). To make inference about the shape of a population distribution, the widely popularregression, would be inadequate, if the distribution is not approximately Gaussian. We present a likelihood-based approach to the estimation of the regression quantiles that uses the asymmetric Laplace density.

온라인 서포트벡터기계를 이용한 온라인 비정상 사건 탐지

박혜정,Park, Hye-Jung 한국데이터정보과학회 2011 한국데이터정보과학회지 Vol.22 No.2

신호처리 관련 응용문제에서는 신호에서 실시간으로 발생하는 비정상적인 사건들을 탐지하는 것이 매우 중요하다. 이전에 알려져 있는 비정상 사건 탐지방법들은 신호에 대한 명확한 통계적인 모형을 가정하고, 비정상적인 신호들은 통계적인 모형의 가정 하에서 비정상적인 사건들로 해석한다. 탐지방법으로 최대우도와 베이즈 추정 이론이 많이 사용되고 있다. 그러나 앞에서 언급한 방법으로는 로버스트 하고 다루기 쉬운 모형을 추정한다는 것은 쉽지가 않다. 좀 더 로버스트한 모형을 추정할 수 있는 방법이 필요하다. 본 논문에서는 로버스트 하다고 알려져 있는 서포트 벡터 기계를 이용하여 온라인으로 비정상적인 신호를 탐지하는 방법을 제안한다. The ability to detect online abnormal events in signals is essential in many real-world signal processing applications. In order to detect abnormal events, previously known algorithms require an explicit signal statistical model, and interpret abnormal events as statistical model abrupt changes. In general, maximum likelihood and Bayesian estimation theory to estimate well as detection methods have been used. However, the above-mentioned methods for robust and tractable model, it is not easy to estimate. More freedom to estimate how the model is needed. In this paper, we investigate a machine learning, descriptor-based approach that does not require a explicit descriptors statistical model, based on support vector machines are known to be robust statistical models and a sequential optimal algorithm online support vector machine is introduced.

서포트벡터 군집분석을 이용한 대구·경북지역 대학의 분류

박혜정,김종태,Park, Hye Jung,Kim, Jong Tae 한국데이터정보과학회 2013 한국데이터정보과학회지 Vol.24 No.4

본 논문에서는 대구 경북지역의 24개 4년제 대학교의 대학공시센터에 공시한 대학지표 자료를 사용하였다. 이들 대학지표들 중 재학생 충원률과 건강보험 취업률에 대한 지표를 이용하여 유사 특징을 가지고 있는 대학들을 그룹화 (분류)하여 그룹의 특징을 분석하는데 목적이 있다. SPSS의 계층적 군집분석과 서포트벡터 군집분석을 분석방법으로 활용하여 실험한 결과에서 공통으로 도출할 수 있는 정보를 구하였다. There are sixteen indicators of "College Information" found on the website of College Information Disclosure Center. Among these indicators, the current study examined an enrollment rate and an employment rate based on health insurance coverage, and focused on twenty-four universities in Daegu and Gyeongbuk area. The universities were classified into groups by the enrollment rate and employment rate. This study investigated the characteristics pertaining to those different groups. Hierarchical cluster analysis and support vector cluster analysis were conducted in order to analyze the characteristics of the groups statistically.

최소제곱 서포트벡터기계를 이용한 시장점유율 자료 분석

박혜정,Park, Hye-Jung 한국데이터정보과학회 2009 한국데이터정보과학회지 Vol.20 No.5

본 논문에서는 시장점유율을 추정할 때 최소제곱 서포트벡터기계를 적용하여 보통최소제곱과 최소제곱 서포트벡터기계의 성능을 비교하고자 한다. 최소제곱 서포트벡터기계는 커널 함수를 사용함으로 고차원의 특징 공간에서 선형회귀로 재구성함으로 비선형 회귀문제까지도 해결할 수 있는 장점을 가지고 있다. 그래서 본 논문에서는 비모수 기법인 최소제곱 서포트벡터기계를 이용하여 시장점유율 모형을 추정하고자 한다. 최소제곱 서포트벡터기계를 기반으로 한 모형 추정은 시장점유율 유인모형을 해결하기 위한 좋은 대안이 된다. 최소제곱 서포트벡터기계의 성능을 평가하기 위해 비교 실험에서는 한국 자동차 시장에서 차량 판매량을 이용하여 브랜드별 시장점유율 모형을 추정하였다. The purpose of this article is to present the application of Least Squares Support Vector Machine in analyzing the existing structure of brand. We estimate the parameters of the Market Share Attraction Model using a non-parametric technique for function estimation called Least Squares Support Vector Machine, which allows us to perform even nonlinear regression by constructing a linear regression function in a high dimensional feature space. Estimation by Least Squares Support Vector Machine technique makes it a good candidate for solving the Market Share Attraction Model. To illustrate the performance of the proposed method, we use the car sales data in South Korea's car market.

인터넷 구매의도의 선행변수 - 충동구매, 마켓메이븐 성향, 온라인 상호작용 준비성, 온라인 소비자 연기 -

박혜정,Park, Hye-Jung 복식문화학회 2017 服飾文化硏究 Vol.25 No.5

In order to identify the antecedents of internet purchasing intention toward fashion items, this study examines shopping-related variables as both direct antecedents of internet purchasing intention, and as indirect antecedents of internet purchasing intention through online-related variables. Impulse buying and market mavenism were considered as shopping-related variables, whilst online interaction readiness and online consumer procrastination were considered as online-related variables. It was hypothesized that impulse buying and market mavenism not only directly influence purchasing intention toward fashion items, but also indirectly influence it through online interaction readiness and online consumer procrastination. Data were gathered by surveying university students in Seoul using convenience sampling. A total of 286 questionnaires were used in the statistical analysis. SPSS was used for exploratory factor analysis, and AMOS was used for confirmatory factor analysis and path analysis. The factor analysis of market mavenism, impulse buying, and online consumer procrastination revealed one dimension, whilst the factor analysis of online interaction readiness revealed two dimensions: 'online relationship' and 'internet role.' Tests of the hypothesized path proved that impulse buying indirectly influences internet shopping intention only through online consumer procrastination, whereas market mavenism influences internet shopping intention indirectly through both online interaction readiness and online consumer procrastination. The results will be useful for Internet shopping mall marketers and for future study.

고상법으로 합성한 LiMn₂O₄:Al의 전기화학적 특성

박혜정,박선민,노광철,한정화,Park, Hye-Jung,Park, Sun-Min,Roh, Gwang-Chul,Han, Cheong-Hwa 한국세라믹학회 2011 한국세라믹학회지 Vol.48 No.6

Al doped $LiMn_2O_4$ ($LiMn_2O_4:Al$) synthesized by several Al doping process and Solid State method. The Al contents in $Mn_{1-x}Al_xO_2$ for $LiMn_2O_4:Al$ were analyzed 1.7 wt% by EDS. The $LiMn_2O_4:Al$ confirmed cubic spinel structure and approximately 5 ${\mu}m$ particles regardless of three kinds of doping process by solid state method. In the result of electrochemical performances, initial discharge capacity had 115 mAh/g in case of $LiMn_2O_4$ and 111 mAh/g of $LiMn_2O_4:Al$ after 100th cycle at room temperature. But the capacity retention results showed that $LiMn_2O_4$ and $LiMn_2O_4:Al$ were 44% and 69% respectively in the 100th cycle at 60$^{\circ}C$. Therefore we are confirmed that $LiMn_2O_4:Al$ increased the capacity retention about 25% than $LiMn_2O_4$, thus the effect of Al dopping on $LiMn_2O_4$ capacity retention.

편측성 불완전 소구순열의 치료

박혜정,정휘동,정영수,Park, Hye Jeong,Jung, Hwi-Dong,Mulliken, John Butler,Jung, Young-Soo 대한악안면성형재건외과학회 2013 Maxillofacial Plastic Reconstructive Surgery Vol.35 No.3

Cleft lip arises from congenital underdevelopment with various degrees and patterns. Mulliken named a unilateral incomplete cleft lip with no severe cleft as a lesser-form cleft lip and categorized it into three subgroups. Anatomically categorized subgroups are minor-form, microform, and mini-microform cleft by the extent of vermilion-cutaneous dysjuntion. The vermillion cutaneous notch is more than 3 mm from the regular Cupid's bow peak for minor-form, less than 3 mm for microform, and almost no gap with discontinuity on the vermillion cutaneous border for mini-microform. The treatments are rotational advancement flap for minor-form, double unillimb Z-plasty for microform, and vertical lenticular excision for mini-microform, respectively. This article aims to present the literature review about the incomplete lesser form cleft lip classified by Mulliken and to report our experiences.

조선후기 사의당의 위치와 건축적 특징에 대한 추론

박혜정,조재모,Park, Hye-Jung,Cho, Jae-Mo 한국건축역사학회 2017 건축역사연구 Vol.26 No.2

The purpose of this study is to estimate the location and space configuration of Sauidang(四宜堂), which is a house in the late Joseon period. Sauidang was built by Hong Man-hoe(洪萬恢), the last son of Princess Jeongmyeong(貞明公主), was used as a house for Pungsan Hong clan(豊山 洪氏) for approximately 150 years, and is now defunct. However, the existence was estimated through Sauidang-ji("四宜堂志") compiled by Gwan-am Hong Gyeong-mo(冠巖 洪敬謨) in 1824. Based on the mention that "Sauidang was located in Hundo-bang(薰陶坊) of Ihyeon-dong(梨峴洞)" and the literature referring to "the site of Myeongryegung Palace(明禮宮)", this study estimated the location of Sauidang. Maps and photographic data in the modern age enabled us to examine the utilization cases of medium- and large-sized lots by examining the correlation with the current Embassy of the People's Republic of China in the Republic of Korea. In addition, it was possible to estimate the spatial configuration of Sauidang with a focus on words to clarify the detailed explanation of Sauidang mentioned in the literature and spatial relationship.

경상북도 일부 고교생을 대상으로 한 도인안교의 스트레스 저하 효과에 대한 파일럿 연구

박혜정,이상재,임병묵,Park, Hye Jung,Lee, Sangjae,Lim, Byungmook 대한예방한의학회 2014 대한예방한의학회지 Vol.18 No.2

Objective : This study was to examine the effects of Doin-Angyo program, evidenced by East Asian Medicine literatures, on stress among smoking adolescents. Method : This study was a prospective community trial using a one-group, pretest-posttest design. As an intervention, Doin-Angyo program was implemented for 10 minutes per session, 3 days a week, for 6 weeks to smoking students who attend a Girl's high school. Psychological stress was measured by self-report questionnaire and physiological stress was measured by salivary cortisol. The final sample for the questionnaire included 18 participants. In addition, pre- and post-salivary cortisol levels of 24 adolescents participating in the last session were compared in order to identify the acute effects of Doin-Angyo program in reducing stress. Results : There were no significant differences in overall stress-test results from the comparisons before and after the intervention. However, we found a significant going-down of the stress level(p = .032) in the question, "About School-work" among six sub-category questions. From the salivary cortisol tests in the last session, we found, in comparison with the standard salivary cortisol density level, $0.3{\mu}g/dL$, the density level tended to go down(p = .062) when higher than the standard, and it went up(p = .001) when lower than the standard, after 10 minute session. Conclusion : The results of this pilot study supported the partial effect of Doin-Angyo program in reducing the stress levels. The study protocol and results can be used to elaborate the community trials design aiming to prove the effect of Korean Medicine based health promotion modalities.

설문 응답에 대한 신뢰도를 반영한 K-척도에 관한 연구

박혜정,피수영,Park, Hye Jung,Pi, Su Young 한국데이터정보과학회 2013 한국데이터정보과학회지 Vol.24 No.1

정보화시대에 인터넷 중독의 심각성은 정보화 사회의 큰 이슈로 부각되고 있다. 인터넷사용이 급증함에 따라 정보화의 역기능도 증가하고 있어 인터넷 중독은 사회적문제로 대두되고 있다. 초고속 무선인터넷 서비스 보급 및 넷북, 스마트 폰 등의 인터넷 접속기기가 더욱 다양화됨에 따라 인터넷 중독 자가진단 검사 척도인 K-척도의 진단기준도 시대변화에 따라 변화가 요구되었으며 2012년 3월에 K-척도의 진단기준이 변경되었다. 본 논문에서는 2012년 변경된 K-척도의 기준으로 경북지역 대학생들의 인터넷 중독 실태와 K-척도 특징들을 살펴보고자 한다. K-척도에서 중독 진단을 위한 조사방식은 응답자가 직접 자신의 중독증상을 주관적 판단에 의해 응답하는 설문방식이므로 응답자의 고의적인 사실 숨김으로 인해 응답오차가 발생할 수 있다. 본 논문에서는 응답오차를 줄이기 위해 변경된 K-척도에 응답자에게 설문 응답에 대한 신뢰할 수 있는 신뢰도 값을 추가적으로 입력하여 분석의 신뢰도를 높이고자 한다. In the Information age, internet addiction has been a big issue in a modern society. The adverse effects of the internet addiction have been increasing at an exponential speed. Along with a great variety of internet-connected device supplies, K-scale diagnostic criteria have been used for the internet addiction self-diagnose tests in the high-speed wireless Internet service, netbooks, and smart phones, etc. The K-scale diagnostic criteria needed to be changed to meet the changing times, and the diagnostic criteria of K-scale was changed in March, 2012. In this paper, we analyze the internet addiction and K-scale features on the actual condition of Gyeongbuk collegiate areas using the revised K-scale diagnostic criteria in 2012. The diagnostic method on internet addiction is measured by the respondents' subjective estimation. Willful error of the respondents can be occurred to hide their truth. In this paper, we add the survey response to the trusted reliability values to reduce response errors on the K-scale on the K-scale, and enhance the reliability of the analysis.