• Journal of the Korean Data and Information Science Society
• /
• 제20권2호
• /
• pp.311-319
• /
• 2009

군집과 분류분석에서 L1 데이터 뎁스를 이용한 DDclust와 DDclass라고 불리는 로버스트한 방법이 Jornsten (2004)에 의하여 제안되었다. SVM-기반방법이 많이 사용되나 이상치가 있는 경우에는 약간의 문제가 있다. 유전자 자료에서는 유전자 수가 많기 때문에 적절한 유전자 선택과정이 필요하다. 따라서 적절한 유전자 또는 유전자 군집을 선택하여 분류에 이용하면 분류의 성능을 향상시킬 수 있다. 이러한 관점에서 뎁스 기반 분류방법과 SVM-기반 분류방법을 비교 연구하여 그 성능을 비교 하였다.

• 응용통계연구
• /
• 제24권5호
• /
• pp.847-859
• /
• 2011

cDNA microarray-based comparative genomic hybridization(CGH) data includes low-intensity spots and thus a statistical strategy is needed to detect subtle differences between different cancer classes. In this study, genes displaying a high frequency of alteration in one of the different classes were selected among the pre-selected genes that show relatively large variations between genes compared to total variations. Utilizing copy-number changes of the selected genes, this study suggests a statistical approach to predict patients' classes with increased performance by pre-classifying patients with similar genetic alteration scores. Two-stage logistic regression model(TLRM) was suggested to pre-classify homogeneous patients and predict patients' classes for cancer prediction; a decision tree(DT) was combined with logistic regression on the set of informative genes. TLRM was constructed in cDNA microarray-based CGH data from the Cancer Metastasis Research Center(CMRC) at Yonsei University; it predicted the patients' clinical diagnoses with perfect matches (except for one patient among the high-risk and low-risk classified patients where the performance of predictions is critical due to the high sensitivity and specificity requirements for clinical treatments. Accuracy validated by leave-one-out cross-validation(LOOCV) was 83.3% while other classification methods of CART and DT performed as comparisons showed worse performances than TLRM.

• 한국인터넷방송통신학회논문지
• /
• 제20권1호
• /
• pp.101-106
• /
• 2020

본 논문은 채널에서 발생되는 부호간 간섭을 줄일 수 있는 QE-MMA 적응 등화 알고리즘의 성능 개선을 위하여 varying stepsize를 적용한 VS-QE-MMA (Varying Stepsize-Quantized Error-MMA)에 관한 것이다. 송신 신호의 고차 통계치와 오차 신호의 부호만을 이용하는 SE-MMA에서 오차 신호의 크기를 power-of-two 연산을 적용하여 탭 계수 갱신시 필요한 승산과 감산을 천이와 감산만으로 대체하여 H/W 응용을 용이하도록 QE-MMA가 등장하였다. QE-MMA는 이와 같이 연산량의 단순화에 의한 적응 등화 성능이 열화되므로 이를 개선하기 위하여 제안 방식인 VS-QE-MMA에서는 적응을 위한 고정 stepsize를 오차 신호의 비선형 변환에 의한 varying stepsize를 적용하였다. 동일한 채널과 신호대 잡음비에서 제안 방식이 기존 QE-MMA보다 개선된 성능을 얻을 수 있음을 시뮬레이션으로 확인하였다. 시뮬레이션 결과 VS-QE-MMA가 QE-MMA보다 모든 성능 지수에서 우월하였으며, 신호대 잡음비가 10dB 이상일 때 varying stepsize의 효과를 얻을 수 있음을 확인하였다.

• 한국수학교육학회지시리즈A:수학교육
• /
• 제55권2호
• /
• pp.171-192
• /
• 2016

As teachers' competency is evaluated based on their teaching performance. pre-service teachers need to have an opportunity to reflect on themselves by systematically analyzing and evaluating their own lesson planning and teaching performance through self-assessment. In this study, we aimed to examine what evaluation criteria for lesson planning and teaching performance pre-service mathematics teachers consider in the process of self-assessment. This study used a mixed-methods research design. To draw the self-evaluation criteria for lesson planning and teaching performance, pre-service self-reported assessments were analyzed using qualitative analyses. In addition, descriptive statistics were used to investigate the pre-service teachers' distribution across the criteria and check the ratio of pre-service mathematics teachers for each element. As a result, it was disclosed that pre-service mathematics teachers considered eight elements in self evaluating their own lesson planning and teaching performance. In addition, we found that pre-service mathematics teachers tended to consider Pedagogical Content Knowledge (PCK) more than Subject-Matter Knowledge (SMK). Moreover, the results of this study provide educational implications for the curriculum in the pre-service teacher's education program.

• 간호행정학회지
• /
• 제22권3호
• /
• pp.251-259
• /
• 2016

Purpose: The purpose of this study was to examine the effects of workplace spirituality and organizational citizenship behavior on nurses' performance of nursing. Methods: This study was done over a 6 month period, with 395 nurses participants from 5 different general hospitals with 400 beds in 4 regions of Korea. Data were gathered using self-reported questionnaires from July 23 to August 21, 2015. Data were analyzed using descriptive statistics, independent t-test, one way ANOVA, Pearson correlation coefficients, and multiple regression analyses with IBM SPSS 21.0 program. Results: Nursing performance showed positive correlations with workplace spirituality and organizational citizenship behavior. The significant predictors of the nursing performance by nurses were a master's degree or higher (${\beta}=.26$), civic virtue (${\beta}=.19$), courtesy (${\beta}=.12$), 40 years of age or older (${\beta}=.09$), a sense of empathy (${\beta}=.07$), and a bachelor's degree in nursing (${\beta}=.06$). These variables explained 39.2% of the variance in nursing performance by nurses. Conclusion: Based on the outcomes of this study, there is a need to develop intervention programs that can inspire nurse to show spontaneous organizational citizenship behavior, and enhance their workplace spirituality. Related variables including academic credentials should also be taken into account.

• 지역사회간호학회지
• /
• 제31권1호
• /
• pp.34-42
• /
• 2020

Purpose: The objective of this study was to identify whether spirituality mediates the relationship between empathy and elderly care performance among Long Term Care (LTC) hospitals nurses in Korea. Methods: The data collection was performed July 1st to August 31th, 2018. Participants were 119 nurses from three long-term care hospitals in Korea. Self-reported questionnaires were administered to assess general characteristics, empathy, spirituality and elderly care performance. The data were analyzed using descriptive statistics, t-test, one-way ANOVA, Pearson's correlation coefficients, and three-stage simple and multiple regression analysis as proposed by Baron and Kenny. Results: The level of elderly care performance of participants was significantly different based on age (F=3.92, p=.010) and nurse's position (t=-2.18, p=.031). Spirituality had a significant mediating effect on the relationship between empathy and elderly care performance (Z=3.64, p<.001). Conclusion: As spirituality completely mediates the relationship between empathy and elderly care performance, it is necessary to develop a nursing education program that applies spirituality and empathy and supports religious activities at an institutional level.

• International Journal of Reliability and Applications
• /
• 제1권2호
• /
• pp.105-113
• /
• 2000

This paper discusses some problems of the TMN-based architecture of integrated network management systems and presents a CORBA-based architecture of distributed network management system with reliability, which maintains location transparency of distributed managed objects and provides an interoperability among heterogeneous networks and systems. In this paper, the proposed distributed management system model has been applied to the TMN based management systems and then considered for qualitative evaluation of system performance. Two different types of implementation models have been proposed: one is a system using a CORBA gateway and the other is a pure CORBA system, which maps all managed objects into CORBA objects. The characteristics of two system model have been compared in the respect of system implementation

• Journal of applied mathematics & informatics
• /
• 제9권2호
• /
• pp.793-806
• /
• 2002

Feature-based similarity retrieval become an important research issue in image database systems. The features of image data are useful to discrimination of images. In this paper, we propose the highspeed k-Nearest Neighbor search algorithm based on Self-Organizing Maps. Self-Organizing Maps (SOM) provides a mapping from high dimensional feature vectors onto a two-dimensional space. The mapping preserves the topology of the feature vectors. The map is called topological feature map. A topological feature map preserves the mutual relations (similarity) in feature spaces of input data. and clusters mutually similar feature vectors in a neighboring nodes. Each node of the topological feature map holds a node vector and similar images that is closest to each node vector. In topological feature map, there are empty nodes in which no image is classified. We experiment on the performance of our algorithm using color feature vectors extracted from images. Promising results have been obtained in experiments.

• ETRI Journal
• /
• 제38권6호
• /
• pp.1207-1217
• /
• 2016

The general method of machine learning has encountered disadvantages in terms of the significant amount of time and effort required for feature extraction and engineering in natural language processing. However, in recent years, these disadvantages have been solved using deep learning. In this paper, we propose a mention pair (MP) model using deep learning, and a system that combines both rule-based and deep learning-based systems using a guided MP as a coreference resolution, which is an information extraction technique. Our experiment results confirm that the proposed deep-learning based coreference resolution system achieves a better level of performance than rule- and statistics-based systems applied separately

• 전자공학회논문지S
• /
• 제34S권5호
• /
• pp.1-14
• /
• 1997

We consider transmission beamforming techniques for frequency-division-duplex (FDD) wireless communication systems using adaptive arrays to improve the signal quality of the array transmission link. We develop a simple effective transmission beamforming technique based on an approximated frequency tranlsation (AFT) to derive the tranmsiion beamforming weights from the uplink channel vector. This technique exploits the invariance of the short-time averaged fast fading statistics to small frequency translations. A simple approximate relationship that relates the transmission channel vector to the reception channel vector is derived. We have developed its practical alternative in which the frequency translation of the channel vector is performed at the principal angle of arrival (AOA) of the u;link synthestic angular spectrum instead of the mean AOA. To analyze the performance of the proposed methods, we consider the power loss incurred by applying the estimated channel vector instead of the true downlink channel vector. The performance is analyzed as a function of the mean AOA, the angular spread, the number of elements, frequncy difference between the uplink and the downlink, and the angle distribution. Their performance is also compared with that of the direct weight reuse method and the AOA based methods.