• Proceedings of the Korea Water Resources Association Conference
• /
• 2017.05a
• /
• pp.135-135
• /
• 2017

최근 강우-유출 모형은 물리적 현상에 근거한 확정론적 모의 모형과 물리적 성분으로 설명할 수 없는 내용에 대해 통계적으로 접근하는 추계학적 모의 모형 등이 계속 연구되고 있어 자연현상에 가까운 결과를 기대할 수 있게 되었다. 하지만 우리나라의 경우 많은 연구에도 불구하고 돌발성 집중호우, 여름철 집중되는 강우 등으로 인해 재난이 반복적으로 발생하고 있어 모형의 정확성에 대한 논의가 지속되고 있다. 동일한 유역에 동일한 입력자료를 사용하더라도 사용하는 모형에 따라 유출 분석결과는 상이하며 이는 유출 해석에 대한 불확실성으로 작용한다. 본 연구에서는 앙상블 및 블랜딩 기법을 사용하여 각 강우-유출 모형의 불확실성을 고려하여 최적 유출량을 산정하고자 한다. 대상 유역으로는 한강 수계에 있는 중랑천 유역을 선정하였으며, Distributed 모형인 Vflo 모형과 Lumped 모형인 저류함수 모형, SSARR모형, TANK 모형을 이용하여 유출 분석을 실시하였다. 그 후, Multi-Model Super Ensemble(MMSE), Simple Model Average(SMA), Mean Square Error(MSE) 방법 등의 blending 기법을 이용하여 하나의 통합된 형태의 유출 분석 결과를 제시하였으며, 최적 유출량 산정을 위한 blending 기법을 선정하였다. 본 연구를 통해 동일한 강우 시나리오에 대한 여러 강우-유출 모형에 대한 정확도를 확인하였으며, 앙상블 및 블랜딩 기법을 사용하여 유출 분석에 대한 정확도를 향상시킬 수 있을 것으로 판단된다.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.17 no.4
• /
• pp.1080-1099
• /
• 2023

Digital healthcare combined with telemedicine services in the form of convergence with digital technology and AI is developing rapidly. Digital healthcare research is being conducted on many conditions including shock. However, the causes of shock are diverse, and the treatment is very complicated, requiring a high level of medical knowledge. In this paper, we propose a shock detection method based on the correlation between shock and data extracted from hemodynamic monitoring equipment. From the various parameters expressed by this equipment, four parameters closely related to patient shock were used as the input data for a machine learning model in order to detect the shock. Using the four parameters as input data, that is, feature values, a random forest-based ensemble machine learning model was constructed. The value of the mean arterial pressure was used as the correct answer value, the so called label value, to detect the patient's shock state. The performance was then compared with the decision tree and logistic regression model using a confusion matrix. The average accuracy of the random forest model was 92.80%, which shows superior performance compared to other models. We look forward to our work playing a role in helping medical staff by making recommendations for the diagnosis and treatment of complex and difficult cases of shock.

• Korean Journal of Agricultural and Forest Meteorology
• /
• v.25 no.3
• /
• pp.129-141
• /
• 2023

Crop models have been used to predict yield under diverse environmental and cultivation conditions, which can be used to support decisions on the management of forage crop. Cultivar parameters are one of required inputs to crop models in order to represent genetic properties for a given forage cultivar. The objectives of this study were to compare calibration and ensemble approaches in order to minimize the uncertainty of crop yield estimates using the SIMPLE crop model. Cultivar parameters were calibrated using Log-likelihood (LL) and Generic Composite Similarity Measure (GCSM) as an objective function for Metropolis-Hastings (MH) algorithm. In total, 20 sets of cultivar parameters were generated for each method. Two types of ensemble approach. First type of ensemble approach was the average of model outputs (E_em), using individual parameters. The second ensemble approach was model output (E_pm) of cultivar parameter obtained by averaging given 20 sets of parameters. Comparison was done for each cultivar and for each error calculation methods. 'Jowoo' and 'Yeongwoo', which are forage rice cultivars used in Korea, were subject to the parameter calibration. Yield data were obtained from experiment fields at Suwon, Jeonju, Naju and I ksan. Data for 2013, 2014 and 2016 were used for parameter calibration. For validation, yield data reported from 2016 to 2018 at Suwon was used. Initial calibration indicated that genetic coefficients obtained by LL were distributed in a narrower range than coefficients obtained by GCSM. A two-sample t-test was performed to compare between different methods of ensemble approaches and no significant difference was found between them. Uncertainty of GCSM can be neutralized by adjusting the acceptance probability. The other ensemble method (E_pm) indicates that the uncertainty can be reduced with less computation using ensemble approach.

• Journal of the Korea Society of Computer and Information
• /
• v.28 no.11
• /
• pp.53-63
• /
• 2023

This study developed a cognitive impairment predictive model as one of the screening tests for preventing dementia in the elderly by using Automated Machine Learning(AutoML). We used 'Wearable lifelog data for high-risk dementia patients' of National Information Society Agency, then conducted using PyCaret 3.0.0 in the Google Colaboratory environment. This study analysis steps are as follows; first, selecting five models demonstrating excellent classification performance for the model development and lifelog data analysis. Next, using ensemble learning to integrate these models and assess their performance. It was found that Voting Classifier, Gradient Boosting Classifier, Extreme Gradient Boosting, Light Gradient Boosting Machine, Extra Trees Classifier, and Random Forest Classifier model showed high predictive performance in that order. This study findings, furthermore, emphasized on the the crucial importance of 'Average respiration per minute during sleep' and 'Average heart rate per minute during sleep' as the most critical feature variables for accurate predictions. Finally, these study results suggest that consideration of the possibility of using machine learning and lifelog as a means to more effectively manage and prevent cognitive impairment in the elderly.

• Journal of Korean Institute of Industrial Engineers
• /
• v.5 no.1
• /
• pp.59-66
• /
• 1979

The objective of this paper is to develop a stochastic inventory system model under the so-called continuous-review policy with a Poisson one-at-a-time demand process, iid customer inter-arrival times {Xi}, backorders allowed, and constant procurement lead time $\gamma$. The distributions of the so-called inventory position process {$IP_{(t-r)}$} and lead time demand process {$D_{(t-r,t)}$} are formulated in terms of cumulative demand by time t, {$N_t$}. Then, for the long-run expected average annual inventory cost expression, the "ensemble" average is estimated, where the cost variations for stock ordering, holding and backorders are considered stationary.

• Journal of Photoscience
• /
• v.11 no.32
• /
• pp.47-53
• /
• 2004

A laser scanning fluorescence microscope system has been fabricated for single molecule detection (SMD). Problems associated with the system set-up have been discussed along with proper suggestions. Based on the SMD results obtained by using the apparatus, a statistical method has been suggested to determine the minimum number of required molecules to form a group of uniform average in a selected error range.

• Journal of the Institute of Electronics Engineers of Korea CI
• /
• v.40 no.1
• /
• pp.19-29
• /
• 2003

This paper presents a novel hybrid classifier for face recognition with artificially generated virtual training samples. We utilize both the nearest neighbor approach in feature angle space and a connectionist model to obtain a synergy effect by combining the results of two heterogeneous classifiers. First, a classifier called the nearest feature angle (NFA), based on angular information, finds the most similar feature to the query from a given training set. Second, a classifier has been developed based on the recall of stored frontal projection of the query feature. It uses a frontal recall network (FRN) that finds the most similar frontal one among the stored frontal feature set. For FRN, we used an ensemble neural network consisting of multiple multiplayer perceptrons (MLPs), each of which is trained independently to enhance generalization capability. Further, both classifiers used the virtual training set generated adaptively, according to the spatial distribution of each person's training samples. Finally, the results of the two classifiers are combined to comprise the best matching class, and a corresponding similarit measure is used to make the final decision. The proposed classifier achieved an average classification rate of 96.33% against a large group of different test sets of images, and its average error rate is 61.5% that of the nearest feature line (NFL) method, and achieves a more robust classification performance.

• Journal of Korea Water Resources Association
• /
• v.44 no.5
• /
• pp.389-406
• /
• 2011

The objective of this study is to examine the climate change impact assessment on Korean water resources considering the uncertainties of Global Climate Models (GCMs) and hydrological models. The 3 different emission scenarios (A2, A1B, B1) and 13 GCMs' results are used to consider the uncertainties of the emission scenario and GCM, while PRMS, SWAT, and SLURP models are employed to consider the effects of hydrological model structures and potential evapotranspiration (PET) computation methods. The 312 ensemble results are provided to 109 mid-size sub-basins over South Korean and Gaussian kernel density functions obtained from their ensemble results are suggested with the ensemble mean and their variabilities of the results. It shows that the summer and winter runoffs are expected to be increased and spring runoff to be decreased for the future 3 periods relative to past 30-year reference period. It also provides that annual average runoff increased over all sub-basins, but the increases in the northern basins including Han River basin are greater than those in the southern basins. Due to the reason that the increase in annual average runoff is mainly caused by the increase in summer runoff and consequently the seasonal runoff variations according to climate change would be severe, the climate change impact on Korean water resources could intensify the difficulties to water resources conservation and management. On the other hand, as regards to the uncertainties, the highest and lowest ones are in winter and summer seasons, respectively.

• Journal of Communications and Networks
• /
• v.10 no.1
• /
• pp.5-9
• /
• 2008

Researchers have investigated many upper bound techniques applicable to error probabilities on the maximum likelihood (ML) decoding performance of turbo-like codes and low density parity check (LDPC) codes in recent years for a long codeword block size. This is because it is trivial for a short codeword block size. Previous research efforts, such as the simple bound technique [20] recently proposed, developed upper bounds for LDPC codes and turbo-like codes using ensemble codes or the uniformly interleaved assumption. This assumption bounds the performance averaged over all ensemble codes or all interleavers. Another previous research effort [21] obtained the upper bound of turbo-like code with a particular interleaver using a truncated union bound which requires information of the minimum Hamming distance and the number of codewords with the minimum Hamming distance. However, it gives the reliable bound only in the region of the error floor where the minimum Hamming distance is dominant, i.e., in the region of high signal-to-noise ratios. Therefore, currently an upper bound on ML decoding performance for turbo-like code with a particular interleaver and LDPC code with a particular parity check matrix cannot be calculated because of heavy complexity so that only average bounds for ensemble codes can be obtained using a uniform interleaver assumption. In this paper, we propose a new bound technique on ML decoding performance for turbo-like code with a particular interleaver and LDPC code with a particular parity check matrix using ML estimated weight distributions and we also show that the practical iterative decoding performance is approximately suboptimal in ML sense because the simulation performance of iterative decoding is worse than the proposed upper bound and no wonder, even worse than ML decoding performance. In order to show this point, we compare the simulation results with the proposed upper bound and previous bounds. The proposed bound technique is based on the simple bound with an approximate weight distribution including several exact smallest distance terms, not with the ensemble distribution or the uniform interleaver assumption. This technique also shows a tighter upper bound than any other previous bound techniques for turbo-like code with a particular interleaver and LDPC code with a particular parity check matrix.

• Transactions of the Korean Society of Mechanical Engineers B
• /
• v.25 no.7
• /
• pp.915-922
• /
• 2001

A new phase-average technique using instantaneous velocity fields obtained by a PIV method has been developed. The technique tracks vorticity centers and estimates the value of circulation for a chosen domain. The locations of vortex-centers and the magnitudes of circulation are matched together then showing a sine wave feature due to the periodic vortex shedding from the circular cylinder. Ensemble averaged and phase averaged velocity fields are successfully measured for the circular cylinder wake where Reynolds number is 3900 based on free stream velocity and cylinder diameter. The convection velocities of the vortices center and the vortex shedding frequency were measured by a single hot-wire probe.