• 한국건축시공학회지
• /
• 제13권2호
• /
• pp.159-168
• /
• 2013

Construction projects have become increasingly complex in recent years, resulting in substantial safety hazards and frequent fall accidents. In an attempt to prevent fall accidents, various safety management systems have been developed. These systems have mainly been evaluated qualitatively and subjectively by practitioners or supervisors, and there are few tools that can be used to quantitatively evaluate the performance of safety management systems. We propose an expertise-based safety performance evaluation model (EXSPEM), which integrates a fuzzy approach-based analytic hierarchy process and a regression approach. The proposed model uses S-shaped curves to represent the degree of contribution by subjective expertise and is verified by a genetic algorithm. To illustrate its practical application, EXSPEM was applied to evaluate the safety performance of a newly developed real-time mobile detector monitoring system. It is expected that this model will be a helpful tool for systematically evaluating the application of a robust safety control and management system in a complex construction environment.

• 한국안전학회지
• /
• 제18권2호
• /
• pp.73-78
• /
• 2003

In order to improve the safety performance oi construction projects, effective and corporative safety management program between head office and job site must be implemented. And its performance must be measured and analyzed for the identification of the problems in the safety management area. This study proposes a previous evaluation model of safety performance for the large construction firm in order to enhance their safety level. The fundamental data for proposed model is based on the past research(Son 2002), which is structured of evaluation criteria. weighted factor. statistical evaluation formula. The model would help the firm management in identifying the weak areas of safety performance in terms of the degree performing certain safety tasks. It is expected that the model could contribute to achieving the "zero accident" level.ot; level.

• 반도체디스플레이기술학회지
• /
• 제21권3호
• /
• pp.63-68
• /
• 2022

In this study, we produced defect detection models using parallel convolution neural networks. If convolution neural networks are constructed parallel type, the model's detection accuracy will increase and detection time will decrease. We produced parallel-type defect detection models using 4 types of convolutional algorithms. The performance of models was evaluated using evaluation indicators. The model's performance is detection accuracy and detection time. We compared the performance of each parallel model. The detection accuracy of the model using AlexNet is 97 % and the detection time is 0.3 seconds. We confirmed that when AlexNet algorithm is constructed parallel type, the model has the highest performance.

• 상하수도학회지
• /
• 제36권4호
• /
• pp.239-248
• /
• 2022

The prediction of algal bloom is an important field of study in algal bloom management, and chlorophyll-a concentration(Chl-a) is commonly used to represent the status of algal bloom. In, recent years advanced machine learning algorithms are increasingly used for the prediction of algal bloom. In this study, XGBoost(XGB), an ensemble machine learning algorithm, was used to develop a model to predict Chl-a in a reservoir. The daily observation of water quality data and climate data was used for the training and testing of the model. In the first step of the study, the input variables were clustered into two groups(low and high value groups) based on the observed value of water temperature(TEMP), total organic carbon concentration(TOC), total nitrogen concentration(TN) and total phosphorus concentration(TP). For each of the four water quality items, two XGB models were developed using only the data in each clustered group(Model 1). The results were compared to the prediction of an XGB model developed by using the entire data before clustering(Model 2). The model performance was evaluated using three indices including root mean squared error-observation standard deviation ratio(RSR). The model performance was improved using Model 1 for TEMP, TN, TP as the RSR of each model was 0.503, 0.477 and 0.493, respectively, while the RSR of Model 2 was 0.521. On the other hand, Model 2 shows better performance than Model 1 for TOC, where the RSR was 0.532. Explainable artificial intelligence(XAI) is an ongoing field of research in machine learning study. Shapley value analysis, a novel XAI algorithm, was also used for the quantitative interpretation of the XGB model performance developed in this study.

• 대한임베디드공학회논문지
• /
• 제14권4호
• /
• pp.207-218
• /
• 2019

The purpose of this study is to develop a model that can systematically study the whole learning process of machine learning. Since the existing model describes the learning process with minimum coding, it can learn the progress of machine learning sequentially through the new model, and can visualize each process using the tensor flow. The new model used all of the existing model algorithms and confirmed the importance of the variables that affect the target variable, survival. The used to classification training data into training and verification, and to evaluate the performance of the model with test data. As a result of the final analysis, the ensemble techniques is the all tutorial model showed high performance, and the maximum performance of the model was improved by maximum 5.2% when compared with the existing model using. In future research, it is necessary to construct an environment in which machine learning can be learned regardless of the data preprocessing method and OS that can learn a model that is better than the existing performance.

• 지능정보연구
• /
• 제21권1호
• /
• pp.119-142
• /
• 2015

사용자의 취향과 선호도를 고려하여 정보를 제공하는 추천 시스템의 중요성이 높아졌다. 이를 위해 다양한 기법들이 제안되었는데, 비교적 도메인의 제약이 적은 협업 필터링이 널리 사용되고 있다. 협업 필터링의 한 종류인 모델 기반 협업 필터링은 기계학습이나 데이터 마이닝 모델을 협업 필터링에 접목한 방법이다. 이는 희박성 문제와 확장성 문제 등의 협업 필터링의 근본적인 한계를 개선하지만, 모델 생성 비용이 높고 성능/확장성 트레이드오프가 발생한다는 한계점을 갖는다. 성능/확장성 트레이드오프는 희박성 문제의 일종인 적용범위 감소 문제를 발생시킨다. 또한, 높은 모델 생성 비용은 도메인 환경 변화의 누적으로 인한 성능 불안정의 원인이 된다. 본 연구에서는 이 문제를 해결하기 위해, 군집화 기반 협업 필터링에 마르코프 전이확률모델과 퍼지 군집화의 개념을 접목하여, 적용범위 감소 문제와 성능 불안정성 문제를 해결한 예측적 군집화 기반 협업 필터링 기법을 제안한다. 이 기법은 첫째, 사용자 기호(Preference)의 변화를 추적하여 정적인 모델과 동적인 사용자간의 괴리 해소를 통해 성능 불안정 문제를 개선한다. 둘째, 전이확률과 군집 소속 확률에 기반한 적용범위 확장으로 적용범위 감소 문제를 개선한다. 제안하는 기법의 검증은 각각 성능 불안정성 문제와 확장성/성능 트레이드오프 문제에 대한 강건성(robustness)시험을 통해 이뤄졌다. 제안하는 기법은 기존 기법들에 비해 성능의 향상 폭은 미미하다. 또한 데이터의 변동 정도를 나타내는 지표인 표준 편차의 측면에서도 의미 있는 개선을 보이지 못하였다. 하지만, 성능의 변동 폭을 나타내는 범위의 측면에서는 기존 기법들에 비해 개선을 보였다. 첫 번째 실험에서는 모델 생성 전후의 성능 변동폭에서 51.31%의 개선을, 두 번째 실험에서는 군집 수 변화에 따른 성능 변동폭에서 36.05%의 개선을 보였다. 이는 제안하는 기법이 성능의 향상을 보여주지는 못하지만, 성능 안정성의 측면에서는 기존의 기법들을 개선하고 있음을 의미한다.

• Asian Pacific Journal of Cancer Prevention
• /
• 제16권8호
• /
• pp.3429-3434
• /
• 2015

Background: This study analyzed predictors of regular mammography performance in Korea. In addition, we determined factors affecting regular mammography performance in life-transition aged women by applying an attitude, social influence, and self-efficacy (ASE) model. Materials and Methods: Data were collected from women aged over 40 years residing in province J in Korea. The 178 enrolled subjects provided informed voluntary consent prior to completing a structural questionnaire. Results: The overall regular mammography performance rate of the subjects was 41.6%. Older age, city residency, high income and part-time job were associated with a high regular mammography performance. Among women who had undergone more breast self-examinations (BSE) or more doctors' physical examinations (PE), there were higher regular mammography performance rates. All three ASE model factors were significantly associated with regular mammography performance. Women with a high level of positive ASE values had a significantly high regular mammography performance rate. Within the ASE model, self-efficacy and social influence were particularly important. Logistic regression analysis explained 34.7% of regular mammography performance and PE experience (${\beta}=4.645$, p=.003), part-time job (${\beta}=4.010$, p=.050), self-efficacy (${\beta}=1.820$, p=.026) and social influence (${\beta}=1.509$, p=.038) were significant factors. Conclusions: Promotional strategies that could improve self-efficacy, reinforce social influence and reduce geographical, time and financial barriers are needed to increase the regular mammography performance rate in life-transition aged.

• 한국추진공학회:학술대회논문집
• /
• 한국추진공학회 2004년도 제22회 춘계학술대회논문집
• /
• pp.513-520
• /
• 2004

A steady-state/transient performance simulation model was newly developed for the propulsion system of the CRW (Canard Rotor Wing) type UAV (Unmanned Aerial Vehicle) during flight mode transition. The CRW type UAV has a new concept RPV (Remotely Piloted Vehicle) which can fly at two flight modes such as the take-off/landing and low speed forward flight mode using the rotary wing driven by engine bypass exhaust gas and the high speed forward flight mode using the stopped wing and main engine thrust. The propulsion system of the CRW type UAV consists of the main engine system and the duct system. The flight vehicle may generally select a proper type and specific engine with acceptable thrust level to meet the flight mission in the propulsion system design phase. In this study, a turbojet engine with one spool was selected by decision of the vehicle system designer, and the duct system is composed of main duct, rotor duct, master valve, rotor tip-jet nozzles, and variable area main nozzle. In order to establish the safe flight mode transition region of the propulsion system, steady-state and transient performance simulation should be needed. Using this simulation model, the optimal fuel flow schedules were obtained to keep the proper surge margin and the turbine inlet temperature limitation through steady-state and transient performance estimation. Furthermore, these analysis results will be used to the control optimization of the propulsion system, later. In the transient performance model, ICV (Inter-Component Volume) model was used. The performance analysis using the developed models was performed at various flight conditions and fuel flow schedules, and these results could set the safe flight mode transition region to satisfy the turbine inlet temperature overshoot limitation as well as the compressor surge margin. Because the engine performance simulation results without the duct system were well agreed with the engine manufacturer's data and the analysis results using a commercial program, it was confirmed that the validity of the proposed performance model was verified. However, the propulsion system performance model including the duct system will be compared with experimental measuring data, later.

• 한국태양에너지학회 논문집
• /
• 제27권3호
• /
• pp.135-142
• /
• 2007

The model to predict flow duration characteristics and performance for small scale hydro power(SSHP) plants is studied to analyze the effects of rainfall condition. One existing SSHP plant was selected and performance characteristics was analyzed by using the developed model. The predicted results from the model developed show that the data were in good agreement with operational results of existing SSHP plant. The results show that both the scale parameter and the shape parameter have large effects on the performance of SSHP sites. And also it was found that the model developed in this study can be a useful tool to predict the performance of SSHP sites.

• 품질경영학회지
• /
• 제33권4호
• /
• pp.42-54
• /
• 2005

This paper has the purpose to find out the causal model between QCC activities and performance. This study consists of four hypotheses. First, QCC teamwork has positive influence on QCC performance. Second, QCC atmosphere has positive influence on QCC performance. Third, QCC autonomy has positive influence on QCC performance. Fourth, this causal model is appropriate for representing the relationship between QCC activities and performance. The results of hypothesis testing are as follows. The first and the fourth hypotheses are adopted. The second and the third hypotheses are rejected. Therefore, QCC teamwork will accelerate QCC activities more than atmosphere and autonomy of QCC.