http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
영아 자녀를 둔 어머니의 양육스트레스 변화가 온정적 양육행동 및 유아의 사회정서발달에 미치는 영향
박새롬(Saerom Park),박혜준(Hye Jun Park) 한국아동학회 2016 아동학회지 Vol.37 No.5
Objective: This study aimed to investigate the trajectory of maternal parenting stress during infancy according to the employment status of mothers when each year`s depression was controlled and examine the longitudinal effects of parenting stress on affective parenting and socioemotional developmental outcomes for preschoolers. Methods: Using latent growth modeling, the first 4 waves of panel data collected from 2,078 mothers and their babies were analyzed. This nationwide data was provided from the 2008 Panel Study on Korean Children, conducted by the Korea Institute of Child Care and Education. Results: First, maternal parenting stress showed a linear increasing pattern, and there were individual differences in intercept and slope. Second, when each year`s maternal depression was controlled, the intercept of parenting stress was lower, but the rate of increase was higher. In multiple group analyses, this increasing pattern of parenting stress was significant only in unemployed mothers. Lastly, the increase of unemployed mothers` parenting stress had direct effects on child emotionality and peer interaction and had an indirect effect on child peer interaction (mediated by affective parenting). Conclusion: The current study verified that the longitudinal effects of increasing parenting stress of unemployed mothers on affective parenting and child socioemotional development when maternal depression level was controlled. These results imply that we need to pay more attention to emotional health and resources of unemployed mothers and provide them with social support as well as employed mothers.
라벨 임베딩 분포를 사용한 효율적인 딥러닝 앙상블 방법
박새롬(Saerom Park) 한국컴퓨터정보학회 2021 韓國컴퓨터情報學會論文誌 Vol.26 No.1
본 연구에서는 레이블 임베딩의 분포를 반영하는 딥러닝 모형을 위한 새로운 스태킹 앙상블 방법론을 제안하였다. 제안된 앙상블 방법론은 기본 딥러닝 분류기를 학습하는 과정과 학습된 모형으로 부터 얻어진 레이블 임베딩을 이용한 군집화 결과로부터 소분류기들을 학습하는 과정으로 이루어져 있다. 본 방법론은 주어진 다중 분류 문제를 군집화 결과를 활용하여 소 문제들로 나누는 것을 기본으로 한다. 군집화에 사용되는 레이블 임베딩은 처음 학습한 기본 딥러닝 분류기의 마지막 층의 가중치로부터 얻어질 수 있다. 군집화 결과를 기반으로 군집화 내의 클래스들을 분류하는 소분류기들을 군집의 수만큼 구축하여 학습한다. 실험 결과 기본 분류기로부터의 레이블 임베딩이 클래스 간의 관계를 잘 반영한다는 것을 확인하였고, 이를 기반으로 한 앙상블 방법론이 CIFAR 100 데이터에 대해서 분류 성능을 향상시킬 수 있다는 것을 확인할 수 있었다. In this paper, we propose a new stacking ensemble framework for deep learning models which reflects the distribution of label embeddings. Our ensemble framework consists of two phases: training the baseline deep learning classifier, and training the sub-classifiers based on the clustering results of label embeddings. Our framework aims to divide a multi-class classification problem into small sub-problems based on the clustering results. The clustering is conducted on the label embeddings obtained from the weight of the last layer of the baseline classifier. After clustering, sub-classifiers are constructed to classify the sub-classes in each cluster. From the experimental results, we found that the label embeddings well reflect the relationships between classification labels, and our ensemble framework can improve the classification performance on a CIFAR 100 dataset.
잠재프로파일 분석을 통한 어머니-유아 상호작용 유형 분류와 예측요인 탐색, 유형별 유아의 인지 및 사회정서발달
박새롬 ( Park Saerom ) 열린부모교육학회 2022 열린부모교육연구 Vol.14 No.3
본 연구의 목적은 만 5세 유아-어머니 상호작용의 질과 빈도에 따라 상호작용 유형을 구분하고, 이러한 유형의 분류를 유아의 기질, 어머니의 심리적 건강 및 양육 관련 요인이 예측할 수 있는지 확인하며, 상호작용 유형에 따라 유아의 인지 및 사회정서발달 특성에 차이가 있는지를 밝히는 것이었다. 이를 위해 한국아동패널의 5-7차 연도에 수집된 1,079가구의 자료에 대하여 잠재프로파일 분석, 다항 로지스틱 회귀분석, 그리고 일원변량분석을 실시하였다. 분석 결과 첫째, 유아 어머니의 상호작용 유형은 온정/관여형, 온정/평균형, 통제/제한형의 3개 유형으로 분류되었다. 둘째, 유아 성별과 가족 사회인구학적 변인의 영향력을 제거하였을 때 유아의 활동성 기질, 어머니의 자아존중감과 양육스트레스, 아버지의 양육참여 수준이 어머니의 상호작용 유형을 예측하는 요인으로 나타났다. 셋째, 세 가지 상호작용 유형 중 온정/관여형의 경우 다른 유형보다 이후 유아의 학업능력 및 사회적 유능감 발달 수준이 더 높고 외현화문제는 더 낮은 것으로 나타났다. 본 연구는 유아기 어머니의 상호작용 유형이 상호작용의 질과 빈도에 따라 서로 다른 유형으로 구분되며 유아의 전반적 발달과 관련이 있음을 밝히고, 어머니의 심리적 건강을 지원하는 부모교육과 아버지의 양육참여를 증진하는 정책적 지원의 필요성을 실증적으로 확인하였다는 의의가 있다. This study aims to investigate the types of mother-child interaction according to the quality and frequency of interaction, identify the effects of child temperament, maternal psychological factors, and parenting-related factors on the different types, and analyze differences in cognitive and socio-emotional development of children by the type of interaction. Latent profile analysis, multinominal logistic regression, and one-way ANOVA were conducted using data on 1,079 households from the Panel Study on Korean Children collected from 2012(5th wave) to 2014(7th wave). The main results are as follows. First, the latent profiles of mother-child interaction were classified into three types: warm/high involvement, warm/average involvement, and control/low involvement. Second, when the influence of the child's gender and socio-demographic characteristics were eliminated, the activity temperament of the child, maternal self-esteem and parenting stress, and paternal parenting participation predicted the mother-child interaction type. Third, children of mothers in the warm/high involvement type had higher academic skills and social competence and lower externalization problems than other types. Revealed differences in the overall development of preschoolers according to the type of mother-child interaction, this study empirically confirmed the need for parent education to support maternal psychological health and policy to encourage fathers to participate in parenting.
지하수 내 과불화화합물 제거를 위한 영가철 기반 환원 기술 개발
박새롬 ( Saerom Park ) 한국수처리학회(구 한국수처리기술연구회) 2020 한국수처리학회지 Vol.28 No.5
Perfluoroalkyl acids have been used as components of aqueous fire-fighting foams (AFFFs). Fire-fighting training with AFFFs at military sites has resulted in the contamination of groundwater with perfluorooctonoic acid (PFOA) and perfluorooctanesulfonate (PFOS) at concentrations higher than the US EPA Provisional Health Advisory levels. PFOA and PFOS are resistant to microbial degradation as well as most conventional abiotic techniques. Low heat-activated persulfate oxidation of PFOA has been demonstrated successfully; however, PFOS remained unaltered even at high heat applied. Given the high oxidation state of carbon in the C-S bond of PFOS, reductive technologies are feasible. Therefore, this study assessed the effectiveness of nanoscale zero valent iron (Fe°) coated with palladium (Pd°) at different conditions (mass, intial pH, and heat) and addition of aluminum (Al°) or clay as a another reductant or supporting material, respectively were evaluated. Addition of Pd catalyst enhanced PFOS removal in Pdo/Feo NPs alone system. And use of clay in Pdo/Feo NPs systems also increased PFOA and PFOS removal up to 75% and 95%, respectively. Also, addition of Al increased PFOS removal at acidic pH condition. However, most of PFOA and PFOS removal observed in all assessed systems were due to strong adsorption onto metals not their destruction which were supported by lack of F<sup>-</sup> and SO<sub>4</sub><sup>2-</sup> measurement as products of their destruction.