• 한국정보통신학회논문지
• /
• 제26권1호
• /
• pp.70-75
• /
• 2022

위치를 정확하게 측정하는 것은 다양한 서비스를 제공하는 데 필요하다. 실내 측위를 위한 데이터는 스마트 폰의 앱을 통해 WiFi 장치로부터 RSSI 값을 측정한다. 이렇게 측정된 데이터는 기계학습의 원시 데이터가 된다. 특징 데이터는 측정된 RSSI 값이고, 레이블은 측정한 위치에 대한 공간의 이름으로 한다. 이를 위한 기계학습 기법은 분류에 효율적인 기법을 적용하여 WiFi 신호만으로 정확한 위치를 예측하는 기법을 연구하고자 한다. 앙상블은 하나의 모델보다 다양한 모델을 통하여 더 정확한 예측값을 구하는 기법으로, bagging과 boosting이 있다. 이 중 Boosting은 샘플링한 데이터를 바탕으로 모델링한 결과를 통해 모델의 가중치를 조정하는 기법으로, 다양한 알고리즘이 있다. 본 연구는 위 기법 중 XGboost를 이용하고, 다른 앙상블 기법과 이용한 수행결과를 바탕으로 성능을 평가한다.

• 유체기계공업학회:학술대회논문집
• /
• 유체기계공업학회 2002년도 유체기계 연구개발 발표회 논문집
• /
• pp.273-278
• /
• 2002

The characteristics of tip vortex within a blade tip region were examined experimentally in various flow coefficients by the way of changing tip clearance and blade stagger angle in an axial Low Speed Research Compressor(LSRC). The objective was to identify the unsteady pressure distribution in the blade passage by ensemble average technique acquired from high-frequency response pressure transducers and the tip vortex by root mean square value(RMS value). Data were reduced statistically using phase-lock technique for detailed pressure distributions.

• International Journal of Vascular Biomedical Engineering
• /
• 제1권1호
• /
• pp.24-31
• /
• 2003

As endothelial cells are subject to flow shear stress, it is important to determine the detailed velocity distribution in microvessels in the study of mechanical interactions between blood and endothelium. This paper describes a velocity field of the arteriole in the rat mesentery using an intravital microscope and high-speed digital video system obtained by a highly accurate PIV technique. Red blood cells (RBCs) velocity distributions with spatial resolutions of $0.8{\times}0.8{\mu}m$ were obtained even near the wall in the center plane of the arteriole. By making ensemble-averaged time-series of velocity distributions, velocity profiles over different cross-sections were calculated for comparison. The shear rate at the vascular wall also evaluated on the basis of the ensemble-averaged profiles. It was shown that the velocity profiles were blunt in the center region of the vessel cross-section while they were steep in the near wall region. The wall shear rates were significantly small, compared with those estimated from the Poiseuille profiles.

• 디지털산업정보학회논문지
• /
• 제17권1호
• /
• pp.109-115
• /
• 2021

The number of freshmen at universities is decreasing due to the recent decline in the school-age population, and the survival of many universities is threatened. To overcome this situation, universities are seeking ways to use big data within the school to improve the quality of education. A study on the prediction of dropout students is a representative case of using big data in universities. The dropout prediction can prepare a systematic management plan by identifying students who will drop out of school due to reasons such as dropout or expulsion. In the case of actual on-campus data, a large number of missing values are included because it is collected and managed by various departments. For this reason, it is necessary to construct a model by effectively reflecting the missing values. In this study, we propose a university student dropout prediction model based on eXtreme Gradient Boost that can be applied to data with many missing values and shows high performance. In order to examine the practical applicability of the proposed model, an experiment was performed using data from C University in Chungbuk. As a result of the experiment, the prediction performance of the proposed model was found to be excellent. The management strategy of dropout students can be established through the prediction results of the model proposed in this paper.

• 한국수자원학회논문집
• /
• 제50권1호
• /
• pp.17-28
• /
• 2017

최근 국지성 집중호우 및 돌발홍수와 같은 급격한 기상변화로 인한 기상재해의 발생빈도가 증가함에 따라 고해상도의 기상레이더 강우자료를 사용한 수공학 분야의 연구가 활발하게 진행되고 있다. 기상레이더는 넓은 지역에 걸쳐 실시간으로 강우현상 감시가 가능하며 지상우량계로는 파악이 불가능한 미계측유역을 통과하는 강우장의 이동 및 변동성 파악이 가능한 장점이 있지만 대기 중 존재하는 수상체로부터 반사되는 반사도를 사용하여 강우량을 산정하므로 시공간적 오차가 존재한다. 본 연구에서는 이러한 문제점을 해결하기 위하여 다변량 Copula 함수를 활용하여 레이더 강우에 존재하는 시공간적 오차를 규명하고 레이더 강우앙상블 생산기법을 개발하였다. 개발된 모형으로부터 생산된 레이더 강우앙상블은 통계적 효율기준 분석결과 우수한 모형성능을 확인하였으며 추가적으로 극치호우 및 강우시계열 패턴 분석결과 지상강우의 특성을 효과적으로 재현하는 것을 확인하였다.

• 한국물환경학회지
• /
• 제36권5호
• /
• pp.353-363
• /
• 2020

Hydrologic models can be classified into two types: those for understanding physical processes and those for predicting hydrologic quantities. This study deals with how to use the model to predict today's stream flow based on the system's knowledge of yesterday's state and the model parameters. In this regard, for the model to generate accurate predictions, the uncertainty of the parameters and appropriate estimates of the state variables are required. In this study, a relatively simple hydrologic partitioning model is proposed that can explicitly implement the hydrologic partitioning process, and the posterior distribution of the parameters of the proposed model is estimated using the Markov chain Monte Carlo approach. Further, the application method of the ensemble Kalman filter is proposed for updating the normalized soil moisture, which is the state variable of the model, by linking the information on the posterior distribution of the parameters and by assimilating the observed steam flow data. The stochastically and recursively estimated stream flows using the data assimilation technique revealed better representation of the observed data than the stream flows predicted using the deterministic model. Therefore, the ensemble Kalman filter in conjunction with the Markov chain Monte Carlo approach could be a reliable and effective method for forecasting daily stream flow, and it could also be a suitable method for routinely updating and monitoring the watershed-averaged soil moisture.

• 정보과학회 논문지
• /
• 제42권9호
• /
• pp.1124-1132
• /
• 2015

일반적으로, 교사 학습 알고리즘이 적절히 학습되기 위해서는 레이블의 편향이 없는 충분한 양의 학습데이터가 필요하다. 그러나 영작문 자동채점 시스템 개발을 위한 충분하고 편향되지 않은 학습데이터를 수집하는 것은 어려운 일이다. 또한 영어 작문 평가의 경우, 전체적인 답안 수준에 대한 다면적인 평가가 이루어진다. 적고 편향되기 쉬운 학습데이터와 이를 이용한 여러 평가영역에 대한 학습모델을 생성해야하기 때문에, 이를 위한 적절한 기계학습 알고리즘을 결정하기 어렵다. 본 논문에서는 이러한 문제를 앙상블학습을 통해 완화할 수 있음을 실험에 통해 보이고자 한다. 실제 중, 고등학교 학생들을 대상으로 시행된 단문형 영작문 채점 결과를 학습데이터 개수와 편향성을 조절하여 실험하였다. 학습데이터의 개수 변화와 편향성 변화의 실험 결과, 에이다부스트 알고리즘을 적용한 결과를 투표로 결합한 앙상블 기법이 다른 알고리즘들 보다 전반적으로 더 나은 성능을 나타냄을 실험을 통해 나타내었다.

• 한국군사과학기술학회지
• /
• 제23권2호
• /
• pp.139-146
• /
• 2020

The camouflage pattern was difficult to distinguish from the surrounding background, so it was difficult to classify the object and the background image when the color image is used as the training data of deep-learning. In this paper, we proposed a red-green color blindness image transformation method using the principle that people of red-green blindness distinguish green color better than ordinary people. Experimental results show that the camouflage soldier's recognition performance improved by proposed a deep learning model of the ensemble technique using the imitated red-green-blind image data and the original color image data.

• International Journal of Advanced Culture Technology
• /
• 제9권4호
• /
• pp.268-273
• /
• 2021

In this paper, we propose a sentiment analysis model that improves performance on small-scale data. A sentiment analysis model for small-scale data is proposed and verified through experiments. To this end, we propose Bagging-Bi-GRU, which combines Bi-GRU, which learns GRU, which is a variant of LSTM (Long Short-Term Memory) with excellent performance on sequential data, in both directions and the bagging technique, which is one of the ensembles learning methods. In order to verify the performance of the proposed model, it is applied to small-scale data and large-scale data. And by comparing and analyzing it with the existing machine learning algorithm, Bi-GRU, it shows that the performance of the proposed model is improved not only for small data but also for large data.

• 한국물환경학회지
• /
• 제27권3호
• /
• pp.264-272
• /
• 2011

The objective of this study was to develop a model for predicting long-term runoff in a basin using the ensemble streamflow prediction (ESP) technique and review its reliability. To achieve the objective, this study improved not only the ESP technique based on the ensemble scenario analysis of historical rainfall data but also conventional ESP techniques used in conjunction with qualitative climate forecasting information, and analyzed and assessed their improvement effects. The model was applied to the Geum River basin. To undertake runoff forecasting, this study tried three cases (case 1: Climate Outlook + ESP, case 2: ESP probability through monthly measured discharge, case 3: Season ESP probability of case 2) according to techniques used to calculate ESP probabilities. As a result, the mean absolute error of runoff forecasts for case 1 proposed by this study was calculated as 295.8 MCM. This suggests that case 1 showed higher reliability in runoff forecasting than case 2 (324 MCM) and case 3 (473.1 MCM). In a discrepancy-ratio accuracy analysis, the Climate Outlook + ESP technique displayed 50.0%. This suggests that runoff forecasting using the Climate Outlook +ESP technique with the lowest absolute error was more reliable than other two cases.