http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
Two-Dimensional Attention-Based LSTM Model for Stock Index Prediction
Yeonguk Yu,김윤중 한국정보처리학회 2019 Journal of information processing systems Vol.15 No.5
This paper presents a two-dimensional attention-based long short-memory (2D-ALSTM) model for stockindex prediction, incorporating input attention and temporal attention mechanisms for weighting of importantstocks and important time steps, respectively. The proposed model is designed to overcome the long-termdependency, stock selection, and stock volatility delay problems that negatively affect existing models. The 2DALSTMmodel is validated in a comparative experiment involving the two attention-based models multi-inputLSTM (MI-LSTM) and dual-stage attention-based recurrent neural network (DARNN), with real stock databeing used for training and evaluation. The model achieves superior performance compared to MI-LSTM andDARNN for stock index prediction on a KOSPI100 dataset.
수압파쇄시험 해석을 위한 중공원통 인장시험과 압열인장시험 화강암 인장강도 비교
조영욱(Yeonguk Jo),장찬동(Chandong Chang),이태종(Tae Jong Lee),김광염(Kwang-Yeom Kim) 한국암반공학회 2013 터널과지하공간 Vol.23 No.5
수압파쇄법으로 최대수평주응력 크기 규명에 필요한 요소 중 하나인 암반의 인장강도를 측정하는 방법에 대해 연구하였다. 석모도 시추공에서 회수한 화강암 시료에 대해 두 가지 실내시험(중공원통 인장시험 및 압열인장시험)으로 인장강도를 측정하고 두 결과가 차이를 보이는지 비교하였다. 중공원통 인장시험에서는 높은 수압증가율 상태에서 더 높은 인장강도를 보여, 현장의 수압파쇄시험에서 보인 수압 증가율 상태에서 측정된 인장강도나 그 증가율로 보정된 인장강도를 이용해야한다는 점을 보였다. 인장강도에 대한 수압 증가율 효과와 크기 효과를 보정하면 중공원통 인장시험 결과는 압열인장시험 결과와 유사하게 나타났으며 이는 수압파쇄 인장강도를 위해 압열인장강도를 이용할 수도 있다는 점을 시사한다. We conducted hollow cylinder tensile strength tests and Brazilian tests in Seokmo granite to measure tensile strength necessary for estimating the magnitude of the maximum horizontal principal stress in hydraulic fracturing stress measurements. Two different pressurization rates were used in hollow cylinder tests. Tensile strengths were determined to be higher at higher pressurization rate, which suggests that tensile strength should be measurement at the same rate used in actual in situ hydraulic fracturing tests. Considering the effect of pressurization rate and specimen size on tensile strength, the hollow cylinder tests and Brazilian tests yield similar results each other. This demonstrates that Brazilian tests can be utilized to produce representative tensile strengths for interpretation of hydraulic fracturing test results.
Two-Dimensional Attention-Based LSTM Model for Stock Index Prediction
Yu, Yeonguk,Kim, Yoon-Joong Korea Information Processing Society 2019 Journal of information processing systems Vol.15 No.5
This paper presents a two-dimensional attention-based long short-memory (2D-ALSTM) model for stock index prediction, incorporating input attention and temporal attention mechanisms for weighting of important stocks and important time steps, respectively. The proposed model is designed to overcome the long-term dependency, stock selection, and stock volatility delay problems that negatively affect existing models. The 2D-ALSTM model is validated in a comparative experiment involving the two attention-based models multi-input LSTM (MI-LSTM) and dual-stage attention-based recurrent neural network (DARNN), with real stock data being used for training and evaluation. The model achieves superior performance compared to MI-LSTM and DARNN for stock index prediction on a KOSPI100 dataset.