• Annual Conference on Human and Language Technology
• /
• 2021.10a
• /
• pp.629-632
• /
• 2021

기계 번역(machine translation)은 자연 언어로 된 텍스트를 다른 언어로 자동 번역 하는 기술로, 최근에는 주로 신경망 기계 번역(Neural Machine Translation) 모델에 대한 연구가 진행되었다. 신경망 기계 번역은 일반적으로 자기회귀(autoregressive) 모델을 이용하며 기계 번역에서 좋은 성능을 보이지만, 병렬화할 수 없어 디코딩 속도가 느린 문제가 있다. 비자기회귀(non-autoregressive) 모델은 단어를 독립적으로 생성하며 병렬 계산이 가능해 자기회귀 모델에 비해 디코딩 속도가 상당히 빠른 장점이 있지만, 멀티모달리티(multimodality) 문제가 발생할 수 있다. 본 논문에서는 단어 정렬(word alignment)을 이용한 비자기회귀 신경망 기계 번역 모델을 제안하고, 제안한 모델을 한국어-영어 기계 번역에 적용하여 단어 정렬 정보가 어순이 다른 언어 간의 번역 성능 개선과 멀티모달리티 문제를 완화하는 데 도움이 됨을 보인다.

• Journal of Korean Society of Coastal and Ocean Engineers
• /
• v.23 no.1
• /
• pp.18-25
• /
• 2011

This study presents the damage monitoring method in foundation-structure interface of harbor caisson using vibration-based autoregressive (AR) model. In order to achieve the objective, the following approaches are implemented. Firstly, vibration-based AR model is selected to monitor the damage in foundation-structure interface of caisson structure. Secondly, finite element analysis on a caisson structure model is implemented to evaluate the vibration-based damage monitoring method. Finally, vibration test on a caisson structure model is performed to evaluate applicability of vibration-based AR model method for foundation-structure interface of caisson structure.

• Journal of the Korean Society for Precision Engineering
• /
• v.8 no.4
• /
• pp.76-83
• /
• 1991

자기회귀-이동평균모델에 의하여 시스템의 파라미터를 추정할 수 있는 벡터채널 원형 격자 필터(vector channel circular lattice filter)의 알고리즘을 제시하였다. 이 알고리즘은 스칼라 연산만으로 이루어져 계산이 간단한 장점이 있다. 3자유도 시스템의 시뮬레이션 결과로부터 격자 필터의 성능을 검증하였으며, 1자유도 팔의 고유진동수와 감쇄비를 추정하였다.

• Korean Journal of Culture and Social Issue
• /
• v.16 no.4
• /
• pp.447-468
• /
• 2010

The purpose of this study was to attain a clearer understanding of the reciprocal effects of deviant self-concept and delinquent behaviors by applying a latent state-trait autoregressive modeling approach. Although traditional autoregressive cross-lagged (ARCL) modeling has been widely applied to test the longitudinal reciprocal relationship between the two constructs, it could produce misspecified findings if there were trait-like processes involved in this relationship. The latent state-trait autoregressive(LST-AR) modeling was applied to control trait effects of deviant self-concept and to examine the reciprocal causal relations between the two constructs. Data were taken from a sample of 3,449 eighth graders who were followed annually for 5 years from the Korea Youth Panel Study. The combining LST-AR model with ARCL model substantiated the reciprocal effects of deviant self-concept and delinquent behaviors, even after the stable trait component of deviant self-concept was taken into account. The present findings shed lights on the reciprocal effects of behaviors (i.e., delinquency) and self concepts (i.e., deviant self-concept). Not only did behaviors change corresponding self-concept, but the ways adolescents perceived themselves influenced their behaviors.

• Proceedings of the KSMRM Conference
• /
• 2001.11a
• /
• pp.173-173
• /
• 2001

목적： 1.5T 생체용 자기공명영상기기를 이용한 수소자기공명분광법으로 용액 내 물질의 정량분석에 대한 가능성을 알아보고자 하였다. 대상 및 방법： 0.01％에서 50％까지의 여러 농도를 갖는 포도당+증류수 혼합액의 모델용액을 만들어 생체용 자기공명영상기기와 시험관 nuclear magnetic resonance (NMR) 분광기에서 각각 수소 자기공명분광법을 시행하여 스펙트럼을 얻었다. 또한 12명의 당뇨환자에서 방광내의 소변에 대해 생체용 자기공명영상기기에서 스펙트럼을 얻고 소변을 추출하여 시험관 NMR 분광기에서 수소자기공명분광법을 시행하였다 각각의 방법으로 얻은 스펙트럼 상에서 포도당 농도에 따른 포도당/물 피크의 면적 비의 변화를 구하였고, 통계처리는 상관분석과 단순선형회귀분석을 시행하였고 회귀식을 산출하였다. 또한 생체용 자기공명영상기기를 이용하여 얻은 결과가 객관적인지 알아보기 위해 시험관 NMR 분광기에서 얻은 결과와의 상관관계를 분석하였다.

• Journal of Korean Society of Coastal and Ocean Engineers
• /
• v.29 no.4
• /
• pp.180-188
• /
• 2017

Recently, as the strength of winds and waves increases due to the climate change, abnormal waves such as swells have been also increased, which results in the increase of downtime events of loading/unloading in a harbour. To reduce the downtime events, breakwaters were constructed in a harbour to improve the tranquility. However, it is also important and useful for efficient port operation by predicting accurately and also quickly the downtime events when the harbour operation is in a limiting condition. In this study, numerical simulations were carried out to calculate the wave conditions based on the forecasted wind data in offshore area/outside harbour and also the long-term observation was carried out to obtain the wave data in a harbour. A forecasting method was designed using an auto-regressive (AR) and artificial neural networks (ANN) models in order to establish the relationship between the wave conditions calculated by wave model (SWAN) in offshore area and observed ones in a harbour. To evaluate the applicability of the proposed method, this method was applied to predict wave heights in a harbour and to forecast the downtime events in Pohang New Harbour with highly complex topography were compared. From the verification study, it was observed that the ANN model was more accurate than the AR model.

• KDI Journal of Economic Policy
• /
• v.11 no.3
• /
• pp.39-59
• /
• 1989

본고(本稿)는 Sims가 개발한 방법을 이용하여 우리나라와 같이 경제구조(經濟構造)가 급히 변하는 상황에서의 경제예측(經濟豫測)의 정확도(正確度)를 제고하고자 하는 시도의 일환이다. 본고(本稿)는 예측자의 사전신뢰(事前信賴)를 이용하여 계수의 값에 대하여 사전제약(事前制約)을 부과(賦課)하고 시간변동(時間變動)을 허용하는 변동계수(變動係數)벡타자귀(自歸)(TBVAR)모형(模型)의 추정방법뿐만 아니라 사전제약(事前制約)의 모수(母數)를 선택하는 방법과 오차(誤差)의 분산(分散)이 자기회귀(自己回歸)할 경우의 대처방법 등 예측(豫測)의 정확도(正確度)를 제고시키는 데 실제 사용되는 방법을 설명하고, 6변수모형(變數模型)을 이용하여 TBVAR 모델의 정확도(正確度)를 타(他) 모델과 비교한다. 정부건설(政府建設), 총통화(總通貨), 사채시장이자율(社債市場利子率), 민간건설(民間建設), 실질(實質)GNP 및 소비자(消費者) 물가지수(物價指數) 등 6변수(變數)에 대한 예측의 정확도를 "타일 U"값을 기준으로 비교할 때 TBVAR은 시간변동(時間變動)을 고려하지 않고 사전제약(事前制約)만 적용한 BVAR이나 사전제약(事前制約)도 적용하지 않은 VAR보다 대부분의 변수의 예측에 있어 더 정확하며 민간건설(民間建設)을 제외하고는 OLS보다 예측오차(豫測誤差)가 작게 나타난다.

• Transactions of the Korean Society of Mechanical Engineers
• /
• v.17 no.12
• /
• pp.2949-2961
• /
• 1993

Dynamic cutting process can be represented by a closed-loop0 system consisted of machine tool structure and pure cutting process. On this paper, cutting system is modeled as a six degrees of freedom system using MARV(Modified Autoregressive Vector) model in face milling, and the modeled dynamic cutting process is used to predict dynamic cutting force component. Based on the double modulation principle, a dynamic cutting force model is developed. From the simulated relative displacements between tool and workpiece the dynamic force domponents can be calculated, and the dynamic force can be obtained by superposition of the static force and dynamic force components. The simulated dynamic cutting forces have a good agreement with the measured cutting force.

• Journal of the Institute of Electronics Engineers of Korea SC
• /
• v.38 no.6
• /
• pp.56-64
• /
• 2001

In this paper, we proposed the simple and selective method for the order of model that reflected the feature of the heart rate variability without the complicated calculation in the power spectral analysis of heart rate variability using autoregressive model. The power spectral analysis of short-term of heart rate variability using autoregressive have been problem to resolution of spectral estimates by the selective model order. As a result that the proposed method for the order comparative tested with the AIC and the fixed order method, the calculation process could become very simple and select the order which correspond with the feature of the time series. We verified it could removed the noisy power components by the fixed order.

• The Journal of the Acoustical Society of Korea
• /
• v.18 no.3
• /
• pp.62-67
• /
• 1999

In this paper, we investigates a reliable model of the Predictive Recurrent Neural Network for the speech recognition. Predictive Neural Networks are modeled by syllable units. For the given input syllable, then a model which gives the minimum prediction error is taken as the recognition result. The Predictive Neural Network which has the structure of recurrent network was composed to give the dynamic feature of the speech pattern into the network. We have compared with the recognition ability of the Recurrent Network proposed by Elman and Jordan. ETRI's SAMDORI has been used for the speech DB. In order to find a reliable model of neural networks, the changes of two recognition rates were compared one another in conditions of: (1) changing prediction order and the number of hidden units: and (2) accumulating previous values with self-loop coefficient in its context. The result shows that the optimum prediction order, the number of hidden units, and self-loop coefficient have differently responded according to the structure of neural network used. However, in general, the Jordan's recurrent network shows relatively higher recognition rate than Elman's. The effects of recognition rate on the self-loop coefficient were variable according to the structures of neural network and their values.