http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
Comparative Study of Dimension Reduction Methods for Highly Imbalanced Overlapping Churn Data
Sujee Lee,Bonhyo Koo,Kyu-Hwan Jung 대한산업공학회 2014 Industrial Engineeering & Management Systems Vol.13 No.4
Retention of possible churning customer is one of the most important issues in customer relationship management, so companies try to predict churn customers using their large-scale high-dimensional data. This study focuses on dealing with large data sets by reducing the dimensionality. By using six different dimension reduction methods?Principal Component Analysis (PCA), factor analysis (FA), locally linear embedding (LLE), local tangent space alignment (LTSA), locally preserving projections (LPP), and deep auto-encoder?our experiments apply each dimension reduction method to the training data, build a classification model using the mapped data and then measure the performance using hit rate to compare the dimension reduction methods. In the result, PCA shows good performance despite its simplicity, and the deep auto-encoder gives the best overall performance. These results can be explained by the characteristics of the churn prediction data that is highly correlated and overlapped over the classes. We also proposed a simple out-of-sample extension method for the nonlinear dimension reduction methods, LLE and LTSA, utilizing the characteristic of the data.
Comparative Study of Dimension Reduction Methods for Highly Imbalanced Overlapping Churn Data
Lee, Sujee,Koo, Bonhyo,Jung, Kyu-Hwan Korean Institute of Industrial Engineers 2014 Industrial Engineeering & Management Systems Vol.13 No.4
Retention of possible churning customer is one of the most important issues in customer relationship management, so companies try to predict churn customers using their large-scale high-dimensional data. This study focuses on dealing with large data sets by reducing the dimensionality. By using six different dimension reduction methods-Principal Component Analysis (PCA), factor analysis (FA), locally linear embedding (LLE), local tangent space alignment (LTSA), locally preserving projections (LPP), and deep auto-encoder-our experiments apply each dimension reduction method to the training data, build a classification model using the mapped data and then measure the performance using hit rate to compare the dimension reduction methods. In the result, PCA shows good performance despite its simplicity, and the deep auto-encoder gives the best overall performance. These results can be explained by the characteristics of the churn prediction data that is highly correlated and overlapped over the classes. We also proposed a simple out-of-sample extension method for the nonlinear dimension reduction methods, LLE and LTSA, utilizing the characteristic of the data.
Overstated Expectations on Predicting Stock Markets using Google Trends
Jaewook Lee(이재욱),Han, Sangwoo(한상우),Lee, Sujee(이수지),Koo, Bonhyo(구본효) 대한산업공학회 2014 대한산업공학회 추계학술대회논문집 Vol.2014 No.11
? Human behavior assuredly affects financial markets. What makes it hard to take into consideration those influence is that the behavior of market participants is rather hard to quantify. ? Notwithstanding, data sources, such as Google query volumes or pageviews of Wikipedia, have been suggested as reflections of human behavior. Several studies claimed that stock market fluctuations could be foreseen with such data sets. ? By analyzing Google query volumes, however, we find "early warning signs" works only when stock markets crashes. We also find some patterns may reverse depending on the circumstances.
Ha, Min Woo,Lee, Myungmo,Choi, Sujee,Kim, Seek,Hong, Suckchang,Park, Yohan,Kim, Mi-hyun,Kim, Taek-Soo,Lee, Jihoon,Lee, Jae Kyun,Park, Hyeung-geun American Chemical Society 2015 Journal of organic chemistry Vol.80 No.6
<P>An efficient enantioselective synthetic method for alpha-amido-alpha-alkylmalonates via phase-transfer catalytic alpha-alkylation was successfully developed. The alpha-alkylation of alpha-amidomalonates under phase-transfer catalytic conditions (50% KOH, toluene, -40 degrees C) in the presence of (S,S)-3,4,5-trifluorophenyl-NAS bromide afforded the corresponding alpha-amido-alpha-alkylmalonates in high chemical yields (up to 99%) and optical yields (up to 97% ee), which could be readily converted to versatile chiral intermediates bearing alpha-amino quaternary stereogenic centers. The synthetic potential of this methodology was demonstrated via the synthesis of chiral azlactone, oxazoline, and unnatural alpha-amino acid.</P>
Song, Nan,Sung, Hyuna,Choi, Ji-Yeob,Han, Sohee,Jeon, Sujee,Song, Minkyo,Lee, Yunhee,Park, Chulbum,Park, Sue K,Lee, Kyoung-Mu,Yoo, Keun-Young,Noh, Dong-Young,Ahn, Sei-Hyun,Lee, Sang-Ah,Kang, Daehee American Association for Cancer Research 2012 Cancer Epidemiology, Biomarkers & Prevention Vol.21 No.8
<P>Matrix metalloproteinase-2 (MMP-2) has been thought of as a predictor of recurrence or metastasis risk or prognostic markers in cancer. We evaluated whether preoperative serum levels of MMP-2 work as a prognostic biomarker in breast cancer prognosis.</P>
Combined genetic effect of CDK7 and ESR1 polymorphisms on breast cancer.
Jeon, Sujee,Choi, Ji-Yeob,Lee, Kyoung-Mu,Park, Sue K,Yoo, Keun-Young,Noh, Dong-Young,Ahn, Sei-Hyun,Kang, Daehee M. Nijhoff ; Kluwer Academic Publishers 2010 Breast cancer research and treatment Vol.121 No.3
<P>Breast cancer development is related to genes regulating cell cycle progression such as CCND1, CDK7, and ESR1. We conducted a hospital-based case-control study to evaluate the role of genetic polymorphisms in these genes in breast cancer development among Korean women. Questionnaire data and samples were obtained from 864 incident breast cancer cases and 723 controls recruited from 1995 to 2002. Four single nucleotide polymorphisms (SNPs) in three genes were genotyped by matrix-assisted laser desorption/ionization time-of-flight mass spectrometry [CCND1 Ex4-1G>A (rs9344), CDK7 Ex2-28C>T (rs2972388), ESR1 P325P Ex4-122G>C (rs1801132), and ESR1 T594T Ex8+229G>A (rs2228480)], and their associations with breast cancer risk were assessed. The odds ratios (ORs) and 95% confidence intervals (CIs) were estimated by unconditional logistic regression analysis adjusted for age, education, age at the first full-term pregnancy, and family history of breast cancer. Women carrying the CDK7 TT genotype had increased risk of breast cancer (OR, 1.4; 95% CI, 1.1-1.7). There was no significant association between breast cancer risk and the genetic polymorphisms of CCND1 and ESR1. However, when CDK7 and ESR1 P325P were combined, women carrying both the CDK7 TT and ESR1 P325P CC genotypes showed increased breast cancer risk (OR, 1.7; 95% CI, 1.1-2.5; P for trend, <0.01). In conclusion, our findings suggest that the combined genotypes of CDK7 and ESR1 are associated with breast cancer risk among Korean women.</P>
Learning representative exemplars using one-class Gaussian process regression
Son, Youngdoo,Lee, Sujee,Park, Saerom,Lee, Jaewook Pergamon Press 2018 Pattern Recognition Vol. No.
<P><B>Abstract</B></P> <P>An exemplar is an observation that represents a group of similar observations. Exemplars from data are examined to divide entire heterogeneous data into several homogeneous subgroups, wherein each subgroup is represented by an exemplar. With its inherent sparsity, an exemplar-based learning model provides a parsimonious model to represent or cluster large-scale data. A novel exemplar learning method using one-class Gaussian process (GP) regression is proposed in this study. The proposed method constructs data distribution support from one-class GP regression using automatic relevance determination prior and heterogeneous GP noise. Exemplars that correspond to the basis vectors of the constructed support function are then automatically located during the training process. The proposed method is applied to various data sets to examine its operability, characteristics of data representation, and cluster analysis. The exemplars of some real data generated by the proposed method are also reported.</P> <P><B>Highlights</B></P> <P> <UL> <LI> A novel sparse Bayesian algorithm for learning exemplars is proposed. </LI> <LI> The proposed method automatically locates exemplars among similar observations. </LI> <LI> Applications to data representation and cluster analysis are provided. </LI> <LI> Theoretical generalization error bound for the method is provided. </LI> </UL> </P>
도시와 농어촌 노인의 노화불안과 자기방임: 자녀 지원과 사회적 지원의 조절효과를 중심으로
김수지 ( Kim Sujee ),권은주 ( Kwon Eun Joo ),이장범 ( Lee Jangbum ),김순은 ( Kim Soon Eun ) 한국보건사회연구원 2019 保健社會硏究 Vol.39 No.3
본 연구는 도시 및 농어촌 노인이 인식하는 노화불안, 자기방임행동, 자녀 지원 및 사회적 지원 수준을 알아보고, 노화불안과 자기방임 간 관계에서 자녀의 지원과 사회적 지원의 조절효과를 검증하기 위해 수행되었다. 이를 위해 서울대 SSK고령사회연구단의 [노인의 건강한 노화 및 웰다잉에 관한 연구] 데이터를 활용하여 서울시 및 6개 광역시에 거주하는 도시 노인 837명과 군단위 지역에 거주하는 322명의 농어촌 노인을 비교하였다. 인구학적 특성과 주요 변수에 대한 집단 비교를 위해 t-test 및 χ<sup>2</sup> test를 실시하였고, 회귀분석을 통해 노화불안과 자기방임과의 관계와, 자녀 지원 및 사회적 지원의 조절효과를 살펴보았다. 연구결과, 노화불안 수준에서 집단 간 차이가 발견되었지만, 두 집단 모두 노화불안을 높게 인식할수록 낮은 자기방임 수준을 나타내었다. 자기방임에 대한 자녀 및 사회적 지원 효과는 집단마다 다른 특징을 띄었고, 자녀 지원과 사회적 지원의 조절효과는 농어촌 노인집단에서만 발견되었다. 본 연구는 도농 노인들이 노화이슈에 관련한 스트레스에 대처하고 자기방임행동을 예방하기 위해 자녀 및 사회적 지원을 어떻게 활용하는지에 대해 실증적으로 규명하였다는 점에서 의의가 있다. The goal of this study is to explore aging anxiety, self-neglecting behaviors, levels of family and social support among elders living in urban and rural communities, and to investigate the moderating effects of family and social support on the relationship between aging anxiety and self-neglecting behaviors. Survey data of 837 elders living in urban areas and 322 elders living in rural areas, collected by the Aging Society and Social Capital Research Center in 2018, was analyzed. χ<sup>2</sup> tests and t-tests were used to examine group differences on the levels of aging anxiety, self-neglecting behaviors, and family and social support. The moderating effects of family and social support were tested by multiple regression analyses for each group of urban and rural older adults. The results showed that there was a group difference on the levels of aging anxiety, and that the higher degree of aging anxiety for all the elders in both groups, the lower degree of self-neglecting behaviors. The moderating effects of family and social support were only found among older adults living in rural areas. The significance of this study is to show how urban and rural older adults use family and social support in order to cope with anxiety and stress related to aging issues and to prevent self-neglecting behaviors.