• 한국정보과학회논문지:소프트웨어및응용
• /
• 제34권1호
• /
• pp.11-17
• /
• 2007

패턴 분류에 많이 사용되는 기법 중의 하나인 메모리 기반 추론 알고리즘은 단순히 메모리에 저장된 학습패턴 또는 초월평면과 테스트 패턴간의 거리를 계산하여 가장 가까운 학습패턴의 클래스로 분류하기 때문에 테스트 패턴을 분류하는 기준을 설명할 수 없다는 문제점을 가지고 있다. 이 문제를 해결하기 위하여, 메모리 기반 학습 기법인 RPA를 기반으로 학습패턴들에 내재된 규칙성을 표현하는 IF-THEN 형태의 규칙을 생성하는 점진적 학습 알고리즘을 제안하였다. 하지만, RPA에 의해 생성된 규칙은 주어진 학습패턴 집합에만 충실히 학습되어 overfitting 현상을 보이게 되며, 또한 패턴 공간의 과도한 분할로 인하여 필요 이상으로 많은 개수의 규칙이 생성된다. 따라서, 본 논문에서는 생성된 규칙으로부터 불필요한 조건을 제거함으로써 ovefitting 현상을 해결함과 동시에 생성되는 규칙의 개수를 줄일 수 있는 점진적 규칙 추출 알고리즘을 제안하였으며, UCI Machine Learning Repository의 벤치마크 데이터를 이용하여 제안한 알고리즘의 성능을 입증하였다.

• 한국농공학회논문집
• /
• 제64권1호
• /
• pp.39-50
• /
• 2022

In this study, we suggested the optimal training period for predicting the streamflow using the LSTM (Long Short-Term Memory) model based on the deep learning and CMIP5 (The fifth phase of the Couple Model Intercomparison Project) future climate scenarios. To validate the model performance of LSTM, the Jinan-gun (Seongsan-ri) site was selected in this study. We comfirmed that the LSTM-based streamflow was highly comparable to the measurements during the calibration (2000 to 2002/2014 to 2015) and validation (2003 to 2005/2016 to 2017) periods. Additionally, we compared the LSTM-based streamflow to the SWAT-based output during the calibration (2000~2015) and validation (2016~2019) periods. The results supported that the LSTM model also performed well in simulating streamflow during the long-term period, although small uncertainties exist. Then the SWAT-based daily streamflow was forecasted using the CMIP5 climate scenario forcing data in 2011~2100. We tested and determined the optimal training period for the LSTM model by comparing the LSTM-/SWAT-based streamflow with various scenarios. Note that the SWAT-based streamflow values were assumed as the observation because of no measurements in future (2011~2100). Our results showed that the LSTM-based streamflow was similar to the SWAT-based streamflow when the training data over the 30 years were used. These findings indicated that training periods more than 30 years were required to obtain LSTM-based reliable streamflow forecasts using climate change scenarios.

• 한국전자파학회논문지
• /
• 제19권11호
• /
• pp.1254-1264
• /
• 2008

본 논문에서는 RF 전력 증폭기의 메모리 효과 모델링의 정확성을 향상시키기 위한 확장된 메모리 다항식 모델을 제안하고 검증하였다. 볼테라 커널 중에서 대각행렬의 성분만을 고려하는 기본적인 메모리 다항식 기반의 모델의 정확성을 향상시키기 위하여 지연차수가 다른 성분들에 의한 교차항을 추가하여 확장 모델을 구성하였다. 제안된 확장 메모리 다항식의 복잡성을 메모리리스 모델, 메모리 다항식 모델과 비교하였다. 확장된 모델을 이용하여 비선형 관계식을 행렬식으로 표현한 후, 최소 자승법(least square method)을 이용하여 변수를 추출하는 모델링 기법을 제시하였다. 또한, 제안된 기법과 간접 학습 방식을 이용하여 디지털 사전 왜곡기를 구현하기 위한 디지털 사전 왜곡부 구현 방안 및 디지털 신호 처리(DSP) 방식을 제시하였다. 제안된 모델의 성능을 검증하기 위하여 2.3 GHz 대역의 WiBro 신호를 인가한 10 W급 GaN HEMT 전력 증폭기와 30W급 LDMOS 전력 증폭기에 대하여 모델의 정확도를 비교 검토하였으며, 10W GaN HEMT 전력 증폭기에 대하여 제안된 모델을 이용하는 간접 학습 방식에 기반한 디지털 사전 왜곡기를 적용하여 인접 채널 간섭비(ACPR) 성능을 검증하였다. 제안한 모델은 메모리 다항식에 비하여 모델의 정확성을 향상시키고 10 W GaN HEMT에 대하여 디지털 사전 왜곡기 적용시 기존 방식에 비하여 3차 비선형 영역에서 평균 3 dB의 ACPR 성능 향상을 보여주었다.

• 융합정보논문지
• /
• 제11권11호
• /
• pp.44-50
• /
• 2021

최근 빅데이터를 수용하기 위한 대용량 저장 장치가 필요한 엔터프라이즈 저장 시스템에서는 비용과 크기 대비 직접도가 높은 대용량의 플래시 메모리 기반 저장 장치를 많이 사용하고 있다. 본 논문에서는 엔터프라이즈 대용량 저장 장치의 신뢰도와 이용성에 직접적인 영향을 주는 플래시 메모리 미디어의 수명을 극대화 하기 위해 경사하강법을 적용한 고효율 수명 예측 방법을 제안한다. 이를 위해 본 논문에서는 불량 발생 빈도를 학습하기 위한 메타 데이터를 저장하는 매트릭스의 구조를 제안하고 메타데이터를 이용한 비용 모델을 제안한다. 또한 학습된 범위를 벗어난 불량이 발생 했을 때 예외 상황에서의 수명 예측 정책을 제안한다. 마지막으로 시뮬레이션을 통해 본 논문에서 제안하는 방법이 이전까지 플래시 메모리의 수명 예측을 위해 사용되어 온 고정 횟수 기반 수명 예측 방법과 예비 블록의 남은 비율을 기반으로 하는 수명 예측 방법 대비 수명을 극대화 할 수 있음을 증명하여 우수성을 확인했다.

• 의학교육논단
• /
• 제12권2호
• /
• pp.27-33
• /
• 2010

The purpose of this study was to review the definition of cognitive load (CL), the relationship between CL and instructional design, and to provide a viewpoint of CL in curriculum and instructional design in medical education. Cognitive load theory (CLT) makes use of three hypotheses about the structure of human memory: working memory (WM) is limited in terms of the amount of information it can hold, in contrast with WM, long term memory is assumed to have no limits and organizes information as schemata. CL indicates the mental load on the limitation of WM. CLT has been used to design instructional interventions that help to ease the learning process. Extraneous CL is related to irrelevant instructional interventions, while intrinsic CL is the complexity of the information itself. Germane CL is the cognitive process for acquiring schema formation. It is a necessary CL to achieve deeper comprehension and solve problems. The range of medical education includes complex, multifaceted and knowledge-rich domains with clinical skills and attitudes. Therefore, CLT may be used to guide instructional design in medical education in terms of decreasing extraneous CL, adjusting intrinsic CL and enhancing the germane CL.

• 동의생리병리학회지
• /
• 제30권5호
• /
• pp.347-354
• /
• 2016

This study evaluated the effects of Fructus Corni Officinalis water extract (FCE) on congnitive impairment and Aβ clearance induced by beta amyloid Aβ (1-42) injection in the hippocampus of rat. Aβ (1-42) was injected into the hippocampus using a Hamilton syringe and micropump (5 ㎍/5 ㎕, 1 ㎕/min, each hippocampus bilaterally). FCE was administered orally once a day (100, 250, 500 mg/kg) for 4 weeks after the Aβ (1-42) injection. The acquisition of learning and retention of memory were tested using the Morris water maze. Aβ accumulation and Aβ clearance in the hippocampus were observed using immunostaining. Aβ (1-42) level in plasma was confirmed using enzyme-linked immunosorbent assay (ELISA). FCE significantly shortened the escape latencies during acquisition training trials. FCE significantly increased the number of target heading to the platform site and significantly shortened the time for the 1^sttargetheadingduringtheretentiontesttrial.FCEsignificantlyattenuatedtheAβ accumulation in the hippocampus produced by Aβ (1-42) injection. FCE significantly increased LRP-1 expression around vessels in the hippocampus and Aβ (1-42) levels in plasma. The results suggest that FCE improved cognitive impairment by ameliorate Aβ clearance and Aβ accumulation in the hippocampus. FCE may be a beneficial herbal formulation in treating cognitive impairment including Alzheimer's disease.

• 항공우주시스템공학회지
• /
• 제14권2호
• /
• pp.50-56
• /
• 2020

In this study, we investigated whether long short-term memory (LSTM) can be used in the future to predict F10.7 index data; the F10.7 index is a space environment factor affecting atomic oxygen erosion. Based on this, we compared the prediction performances of LSTM, the Autoregressive integrated moving average (ARIMA) model (which is a traditional statistical prediction model), and the similar pattern searching method used for long-term prediction. The LSTM model yielded superior results compared to the other techniques in the prediction period starting from the max/min points, but presented inferior results in the prediction period including the inflection points. It was found that efficient learning was not achieved, owing to the lack of currently available learning data in the prediction period including the maximum points. To overcome this, we proposed a method to increase the size of the learning samples using the sunspot data and to upgrade the LSTM model.

• 한국초등과학교육학회지:초등과학교육
• /
• 제27권3호
• /
• pp.244-251
• /
• 2008

One of the problems with testing for the presence of $CO_2$ is that the results are not visible. In order to over-come this weak point, a new testing method was developed with BTB indicator equipment, that made the gas visible. After that, the experiment was performed and tested the effects of the new visible equipment. The visible equipment could be adapted to regular class use successfully. Also, it was effective in that it reduced waste of gas, minimized danger potential through use of $CO_2$ canisters and candle-sticks, and also increased knowledge about indicators. The new experimental method and equipment affected the students' interest. It is possible that the students' positive participation was due to their interest in the new apparatus and application of the visual senses. The new system was tested fer its effects on teaching content and helping to produce sustained memory of the content. There were no significant differences between the groups in terms of content learning on initial content memory. However, when students in both groups were tested 4 months later, the visual experiment group sustained memory performance, while the other group showed a significant decrease. Generally, boys score higher than girls in terms of interest and participation in experimental activities. In this case, however, there were no difference between groups. It may have been due to introduction of new equipment and different methods from the textbooks. So, this could increase participation in science using various experiments.

• 한국해안·해양공학회논문집
• /
• 제33권6호
• /
• pp.238-245
• /
• 2021

본 연구에서는 LSTM 모델을 활용하여 낙동강 하구역의 DO 농도 예측을 위한 최적 모델 조건과 적합한 예측변수를 찾기 위한 Case study를 수행하였다. 모델 매개변수 case study 결과, Epoch = 300과 Sequence length = 1에서 상대적으로 높은 정확도를 보였다. 예측변수 case study 결과, DO와 수온을 예측변수로 했을 때 가장 높은 정확도를 보였으며, 이는 DO 농도와 수온의 높은 상관성에 기인한 것으로 판단된다. 상기 결과로부터 낙동강 하구역의 DO 농도 예측에 적합한 LSTM 모델 조건과 예측변수를 찾을 수 있었다.

• ETRI Journal
• /
• 제44권4호
• /
• pp.672-685
• /
• 2022

This article proposes a subject-independent application of brain-computer interfacing (BCI). A 32-channel Electroencephalography (EEG) device is used to measure imagined speech (SI) of four words (sos, stop, medicine, washroom) and one phrase (come-here) across 13 subjects. A deep long short-term memory (LSTM) network has been adopted to recognize the above signals in seven EEG frequency bands individually in nine major regions of the brain. The results show a maximum accuracy of 73.56% and a network prediction time (NPT) of 0.14 s which are superior to other state-of-the-art techniques in the literature. Our analysis reveals that the alpha band can recognize SI better than other EEG frequencies. To reinforce our findings, the above work has been compared by models based on the gated recurrent unit (GRU), convolutional neural network (CNN), and six conventional classifiers. The results show that the LSTM model has 46.86% more average accuracy in the alpha band and 74.54% less average NPT than CNN. The maximum accuracy of GRU was 8.34% less than the LSTM network. Deep networks performed better than traditional classifiers.