• Journal of the Korea Convergence Society
• /
• v.8 no.10
• /
• pp.45-52
• /
• 2017

Drifter is an equipment for observing the characteristics of seawater in the ocean, and it can be used to predict effluent oil diffusion and to observe ocean currents. In this paper, we design models or the prediction of drifter trajectory using machine learning. We propose methods for estimating the trajectory of drifter using support vector regression, radial basis function network, Gaussian process, multilayer perceptron, and recurrent neural network. When the propose mothods were compared with the existing MOHID numerical model, performance was improve on three of the four cases. In particular, LSTM, the best performed method, showed the imporvement by 47.59% Future work will improve the accuracy by weighting using bagging and boosting.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.11 no.10
• /
• pp.1845-1851
• /
• 2007

Background noise and room reverberation are two causes of degradation in speech in listening situations. Many algorithms developed to enhance reverberant speech. In this paper we propose a dereverberation method for enhancement of speech using modified the linear prediction(LP) residual in reverberant room condition. The proposed dereberberation method based on the fact that the signification excitation of the vocal tract system takes place at the instant of glottal closure in voiced speech. Our method used delay information form each sensor, and we need reverberant signals from 3 sensors. We obtain a new LP residual signal using modified IP residual combination which derived form weighting of the LP residual and the Hilbert transform of LP residual. The nature of the coherently added Hilbert envelop has several large amplitude spikes because of the effects of noise and reverberation. This residual of the clean speech is used to excite the time-varying all-pole filter to obtain the enhanced speech. We achieved simulation of proposed algorithm for performance analysis in reverberation environment. The proposed algorithm improves substantially the quality of reverberant speech.

• Proceedings of the Korean Operations and Management Science Society Conference
• /
• 2004.05a
• /
• pp.548-551
• /
• 2004

합리적인 업무 수행도의 예측을 통해 기업은 기존 업무절차의 평가뿐 아니라 업무 개선방안과 새로운 업무의 설계기준을 제시할 수 있다. 본 연구는 업무효율지표들 중 가장 중요한 요소인 업무절차의 수행시간을 예측하는 모형을 제시한다. 일반적으로 기업의 업무는 예측가능하며 장기적으로 안정된 성격을 가진다. 우리는 이러한 특성을 바탕으로 한 대기행렬 모형을 구축하고 그것을 분석하여 정적인 방식의 업무실행 시 수행시간을 예측하였다. 그리고 모형의 성능을 시뮬레이션 기법을 사용하여 평가하였다.

• Journal of the Korea Society for Simulation
• /
• v.30 no.4
• /
• pp.1-8
• /
• 2021

Deep learning is one of the most widely accepted methods for the forecasting of time series data which have the complexity and non-linear behavior. In this paper, we investigate the modification of a state-of-art WaveNet deep learning architecture and walk forward validation (WFV) in order to forecast electric power consumption data 24-hour-ahead. WaveNet originally designed for raw audio uses 1D dilated causal convolution for long-term information. First of all, we propose a modified version of WaveNet which activates real numbers instead of coded integers. Second, this paper provides with the training process with tuning of major hyper-parameters (i.e., input length, batch size, number of WaveNet blocks, dilation rates, and learning rate scheduler). Finally, performance evaluation results show that the prediction methodology based on WFV performs better than on the traditional holdout validation.

• The KIPS Transactions:PartC
• /
• v.12C no.1 s.97
• /
• pp.69-76
• /
• 2005

To effectively use limited resources in wireless cellular networks it is necessary to predict exactly the amount of resources required by handoff calls at a future time. In this paper we propose a method which predicts the amount of resources needed by handoff calls more accurately than the existing method based on Wiener processes. The existing method uses the current demands to predict future demands. Although this method is much simpler than using traffic information from neighbor cells, its prediction error increases as time elapses, leading to waste of wireless resources. By using an exponential parameter to decrease the magnitude of error over time, we show in simulation how to outperform the existing method in resource utilization as well as in prediction of resource demands.

• Korean Journal of Agricultural and Forest Meteorology
• /
• v.25 no.2
• /
• pp.80-98
• /
• 2023

This study explores the effectiveness of various data preprocessing algorithms for improving subseasonal to seasonal (S2S) climate predictions from six climate forecast models and their Multi-Model Ensemble (MME) using a deep learning-based postprocessing model. A pipeline of data transformation algorithms was constructed to convert raw S2S prediction data into the training data processed with several statistical distribution. A dimensionality reduction algorithm for selecting features through rankings of correlation coefficients between the observed and the input data. The training model in the study was designed with TimeDistributed wrapper applied to all convolutional layers of U-Net: The TimeDistributed wrapper allows a U-Net convolutional layer to be directly applied to 5-dimensional time series data while maintaining the time axis of data, but every input should be at least 3D in U-Net. We found that Robust and Standard transformation algorithms are most suitable for improving S2S predictions. The dimensionality reduction based on feature selections did not significantly improve predictions of daily precipitation for six climate models and even worsened predictions of daily maximum and minimum temperatures. While deep learning-based postprocessing was also improved MME S2S precipitation predictions, it did not have a significant effect on temperature predictions, particularly for the lead time of weeks 1 and 2. Further research is needed to develop an optimal deep learning model for improving S2S temperature predictions by testing various models and parameters.

• Proceedings of the Korean Nuclear Society Conference
• /
• 1997.10a
• /
• pp.579-584
• /
• 1997

HP Workstation 및 IBM PC에서 사용 가능한 부수로 해석 코드 MATRA $\alpha$-버전을 개발하였다. 이 코드는 정확도 향상 및 사용자 편리를 위해 COBRA-IV-I코드에 비해 여러가지 기능들이 추가되었으며, 코드의 적용 범위를 신형원자로의 비정방형 집합체 노심에 확장할 수 있도록 압력손실 모형 등이 개선되었다. 또한 이상 유동장에서의 예측 정확도 향상을 위하여 부수로 잔의 횡방향 전달 모형을 개선하였다. 코드의 예측 성능을 평가하기 위해 세 중류의 집합체 유동분포 및 엔탈피 분포 실험 자료와 비교하였으며, 그 결과 기존의 COBRA-IV-I코드보다 향상된 결과를 보였다.

• The Journal of the Acoustical Society of Korea
• /
• v.4 no.1
• /
• pp.43-54
• /
• 1985

이 논문에서는 Frequency weighting을 이용하여 선형예측 부호화기의 명료성을 개선하는 방법 을 연구한다. 잡음이 섞이지 않은 음성에 대해서는 음성을 분석하기전에 frequency weighting을 행한다. 또한 잡음이 섞인 음성인 경우에는 잡음성분을 spectral subtraction 방법에 의해서 제거한 다음에 frequency weighting을 준다. 이 때 frequency weighting을 주기 위해서 귀의 특성과 연관되어 잘 알려 진 C- message weighting 함수, flanagan weighting 함수 및 articulation index를 약간 수정한 weighting 함수를 사용했다. 여러 객관적인 distance measure를 사용하여 frequency weighting 방법의 성능을 측정하고 귀로 들어 본 결과, frequency weighting 방법을 사용하여 선형예측 방법에 의한 합성 음의 명료도를 효율적으로 개선할 수 있었다.

• 한국신재생에너지학회:학술대회논문집
• /
• 2006.11a
• /
• pp.202-205
• /
• 2006

최근 지열원 열펌프 시스템 설치가 해마다 평균 10-30%젓도 꾸주히 증가하고 있다 주요 연구동향은 토양열전도 측정, 지열히트펌프 시스템 전주기 성능평가 하이브리드 시스템의 초기비용 저감과 이러한 지열원 열펌프 시스템 설계방법분야 개발에 대해 초점이 맞춰지고 있다. 특히 국내에서 현재 시공되어진 많은 시스템들이 부실시공의 문제에 노출되고 있으며 이러한 시점에서 현재의 저가 입찰제도 보다는 외국 사례와 같은 성능확인 제도로의 전환 및 많은 연구가 필요하다. 성능확인제도는 사전 성능 예측과 사후 성능 확인 검증으로 구성되며 본 기술현안 보고서는 최근 국내외 연구동향 및 사전 성능 예측과 사후 성능 검증 관하여 정리하여 본다.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.24 no.1
• /
• pp.8-14
• /
• 2020

The human impact of particulate matter are revealed and demand for improved forecast accuracy is increasing. Recently, efforts is made to improve the accuracy of PM10 predictions by using machine learning, but prediction performance is decreasing due to the particulate matter data with a large rate of low concentration occurrence. In this paper, separation prediction model by concentration is proposed to improve the accuracy of PM10 particulate matter forecast. The low and high concentration prediction model was designed using the weather and air pollution factors in Cheonan, and the performance comparison with the prediction models was performed. As a result of experiments with RMSE, MAPE, correlation coefficient, and AQI accuracy, it was confirmed that the predictive performance was improved, and that 20.62% of the AQI high-concentration prediction performance was improved.