• Proceedings of the Korea Water Resources Association Conference
• /
• 2021.06a
• /
• pp.317-317
• /
• 2021

효율적인 물관리를 위한 댐 유입량 대한 연구는 필수적이다. 본 연구에서는 다양한 머신러닝 알고리즘을 통해 40년동안의 기상 및 댐 유입량 데이터를 이용하여 소양강댐 유입량을 예측하였으며, 그 중 고유량과 저유량예측에 적합한 알고리즘을 각각 선정하여 머신러닝 알고리즘을 결합한 CombML을 개발하였다. 의사 결정 트리 (DT), 멀티 레이어 퍼셉트론 (MLP), 랜덤 포레스트(RF), 그래디언트 부스팅 (GB), RNN-LSTM 및 CNN-LSTM 알고리즘이 사용되었으며, 그 중 가장 정확도가 높은 모형과 고유량이 아닌 경우에서 특별히 예측 정확도가 높은 모형을 결합하여 결합 머신러닝 알고리즘 (CombML)을 개발 및 평가하였다. 사용된 알고리즘 중 MLP가 NSE 0.812, RMSE 77.218 m3/s, MAE 29.034 m3/s, R 0.924, R2 0.817로 댐 유입량 예측에서 최상의 결과를 보여주었으며, 댐 유입량이 100 m3/s 이하인 경우 앙상블 모델 (RF, GB) 이 댐 유입 예측에서 MLP보다 더 나은 성능을 보였다. 따라서, 유입량이 100 m3/s 이상 시의 평균 일일 강수량인 16 mm를 기준으로 강수가 16mm 이하인 경우 앙상블 방법 (RF 및 GB)을 사용하고 강수가 16 mm 이상인 경우 MLP를 사용하여 댐 유입을 예측하기 위해 두 가지 복합 머신러닝(CombML) 모델 (RF_MLP 및 GB_MLP)을 개발하였다. 그 결과 RF_MLP에서 NSE 0.857, RMSE 68.417 m3/s, MAE 18.063 m3/s, R 0.927, R2 0.859, GB_MLP의 경우 NSE 0.829, RMSE 73.918 m3/s, MAE 18.093 m3/s, R 0.912, R2 0.831로 CombML이 댐 유입을 가장 정확하게 예측하는 것으로 평가되었다. 본 연구를 통해 하천 유황을 고려한 여러 머신러닝 알고리즘의 결합을 통한 유입량 예측 결과, 알고리즘 결합 시 예측 모형의 정확도가 개선되는 것이 확인되었으며, 이는 추후 효율적인 물관리에 이용될 수 있을 것으로 판단된다.

• Proceedings of the Korean Operations and Management Science Society Conference
• /
• 1997.10a
• /
• pp.173-176
• /
• 1997

In this paper, the classification performances of inductive learning methods are investigated using the credit rating data. The adopted classifiers are Multiple Discriminant Analysis (MDA), C4.5 of Quilan, Multi-Layer Perceptron (MLP) and Cascade Correlation Network (CCN). The data used in this analysis is obtained using the publicly announced rating reports from the three korean rating agencies. The performances of 4 classifiers are analyzed in term of prediction accuracy. The results show that no classifier is dominated by the other classifiers.

• Proceedings of the Korean Operations and Management Science Society Conference
• /
• 1996.04a
• /
• pp.745-748
• /
• 1996

Korea Composite Stock Price Index (KOSPI) was predicted using Cascade Correlation Network (CCN) model. CCN was suggested, by Fahlman and Lebiere [1990], to overcome the limitations of backpropagation algorithm such as step size problem and moving target problem. To test the applicability of CCN as a function approximator to the stock price movements, CCN was used as a tool for univariate time series analysis. The fitting and forecasting performance fo CCN on the KOSPI was compared with those of Multi-Layer Perceptron (MLP).

• Journal of Korean Institute of Industrial Engineers
• /
• v.28 no.2
• /
• pp.147-161
• /
• 2002

In novelty detection, one attempts to discriminate abnormal patterns from normal ones. Novelty detection is quite difficult since, unlike usual two class classification problems, only normal patterns are available for training. Auto-Associative Multi-Layer Perceptron (AAMLP) has been shown to provide a good performance based upon the property that novel patterns usually have larger auto-associative errors. In this paper, we give a mathematical analysis of 2-layer AAMLP's output characteristics and empirical results of 2-layer and 4-layer AAMLPs. Various activation functions such as linear, saturated linear and sigmoid are compared. The 2-layer AAMLPs cannot identify non-linear boundaries while the 4-layer ones can. When the data distribution is multi-modal, then an ensemble of AAMLPs, each of which is trained with pre-clustered data is required. This paper contributes to understanding of AAMLP networks and leads to practical recommendations regarding its use.

• Proceedings of the Acoustical Society of Korea Conference
• /
• 1993.06a
• /
• pp.117-120
• /
• 1993

In this paper, we explore the possibility of URAN(Universally Reconstructable Artificial Neural-network) VLSI chip for speech recognition. URAN, a newly developed analog-digital hybrid neural chip, is discussed in respects to its input, output, and weight accuracy and their relations to its performance on speaker independent digit recognition. Multi-layer perceptron(MLP) nets including a large frame input layer are used to recognize a digit syllable at a forward retrieval. The simulation results using the full and limited floating precision computations for the input, output, and weight variables of the network give the comparable classification performance. An MLP with piecewise linear hidden and output units is also trained successfully using low accuracy computation.

• Proceedings of the KSRS Conference
• /
• 2005.10a
• /
• pp.34-37
• /
• 2005

An efficient and robust neural network-based scheme is introduced in this paper to perform automatic cloud detection. Unlike many existing cloud detection schemes which use thresholding and statistical methods, we used the artificial neural network methods, the multi-layer perceptrons (MLP) with back-propagation algorithm and radial basis function (RBF) networks for cloud detection from Geostationary satellite images. We have used a simple scene (a mixed scene containing only cloud and clear sky). The main results show that the neural networks are able to handle complex atmospheric and meteorological phenomena. The experimental results show that two methods performed well, obtaining a classification accuracy reaching over 90 percent. Moreover, the RBF model is the most effective method for the cloud classification.

• Transactions of the Korean Society for Noise and Vibration Engineering
• /
• v.24 no.9
• /
• pp.675-681
• /
• 2014

This paper proposes a fault diagnosis system for an induction motor. This system uses empirical mode decomposition(EMD) to extract fault signatures and multi-layer perceptron(MLP) neural network to facilitate an accurate fault diagnosis. EMD can not only decompose a signal adaptively but also provide intrinsic mode functions(IMFs) containing natural oscillatory modes of the signal. However, every IMF does not represent fault signature, an IMF selection algorithm based on harmonics and their energy of each IMF is proposed. The selected IMFs are utilized for fault classification using MLP and this system shows approximately 98 % diagnosis accuracy for the fault vibration signal of the induction motor.

• The Transactions of The Korean Institute of Electrical Engineers
• /
• v.66 no.5
• /
• pp.870-876
• /
• 2017

In this paper, we proposed the automatic vision inspection system using multi-layer perceptron to detect the defects occurred on rail surface. The proposed system consists of image acquisition part and analysis part. Rail surface image is acquired as equal interval using line scan camera and lighting. Mean filter and dynamic threshold is used to reduce noise and segment defect area. Various features to characterize the defects are extracted. And they are used to train and distinguish defects by MLP-classifier. The system is installed on HEMU-430X and applied to analyze the rail surface images acquired from Honam-line at high speed up to 300 km/h. Recognition rate is calculated through comparison with manual inspection results.

• Proceedings of the Korean Institute of Intelligent Systems Conference
• /
• 2004.10a
• /
• pp.543-546
• /
• 2004

Bishop과 Nabney에 의해 소개된 기존의 혼합 밀도 네트워크(Mixture Density Network)에서는 조건부 확률밀도 함수의 매개변수들(parameters)이 하나의 MLP(multi-layer perceptron)의 출력 벡터로 주어진다. 최근에는 변형된 혼합 밀도 네트워크(Modified Mixture Density Network)라고 하는 이름으로 조건부 확률밀도 함수의 선분포(priors), 조건부 평균(conditional means), 그리고 공분산(covariances) 등이 각각 독립적인 MLP의 출력벡터로 주어지는 경우를 다룬 연구가 보고된 바 있다. 본 논문에서는 조건부 평균이 입력에 관해 선형인 경우를 위한 버전에 대한 이론과 매트랩 프로그램 개발 및 적용을 다룬다. 본 논문에서는 우선 일반적인 혼합 밀도 네트워크에 대해 간단히 설명하고, 혼합 밀도 네트워크의 출력인 다층 퍼셉트론의 매개변수를 각각 다른 다층 퍼셉트론에서 학습시키는 변형된 혼합 밀도 네트워크를 설명한 후, 각각 다른 다층 퍼셉트론을 통해 매개변수를 얻는 것은 동일하나 평균값은 선형함수를 통해 얻는 혼합 밀도 네트워크 버전을 소개한다. 그리고, 모의실험을 통하여 이러한 혼합 밀도 네트워크를의 적용가능성에 대해 알아본다.

• Proceedings of the Korean Society of Precision Engineering Conference
• /
• 1997.10a
• /
• pp.1032-1035
• /
• 1997

The purpose of this paper is to reduce the number of cutting tests and to predict the main cutting force and the specific cutting energy. By using the SOFM neural network, the most suitable cutting test conditions has been found. As a result, the number of cutting tests has been reduced to one-third. And by using MLP neural network and regression analysis, the main cutting force and specific cutting energy has been predicted. Predicted values of main cutting force and specific cutting energy are well concide with the measured ones.