• The Journal of the Korea Contents Association
• /
• v.9 no.2
• /
• pp.97-105
• /
• 2009

This paper presents a combination of the generalized Cascade Correlation and generalized Recurrent Cascade Correlation learning algorithms. The new network will be able to grow with vertical or horizontal direction and with recurrent or without recurrent units for the quick solution of the pattern classification problem. The proposed algorithm was tested learning capability with the sigmoidal activation function and hyperbolic tangent activation function on the contact lens and balance scale standard benchmark problems. And results are compared with those obtained with Cascade Correlation and Recurrent Cascade Correlation algorithms. By the learning the new network was composed with the minimal number of the created hidden units and shows quick learning speed. Consequently it will be able to improve a learning capability.

• Proceedings of the Korean Institute of Navigation and Port Research Conference
• /
• 2022.11a
• /
• pp.383-384
• /
• 2022

Gas Combustion Unit (GCU) onboard liquefied natural gas carriers handles boil-off to stabilize tank pressure. There are many factors for LNG cargo operators to take into consideration to determine whether to use GCU or not. Gas consumption of main engine and re-liquefied gas through the Partial Re-Liquefaction System (PRS) are good examples of these factors. Human gas operators have decided the operation so far. In this paper, some deep learning neural network models were developed to provide human gas operators with a decision support system. The models consider various factors specially into GCU operation. A deep learning model with Sigmoid activation functions in input layer and hidden layers made the best performance among eight different deep learning models.

• Journal of Sensor Science and Technology
• /
• v.8 no.1
• /
• pp.45-52
• /
• 1999

In this raper, an enhanced learning method is proposed for improving the learning speed of the error back propagation learning algorithm. In order to cope with the premature saturation phenomenon at the initial learning stage, a variation scheme of active functions is introduced by using higher order functions, which does not need much increase of computation load. It naturally changes the learning rate of inter-connection weights to a large value as the derivative of sigmoid function abnormally decrease to a small value during the learning epoch. Also, we suggest the hybrid learning method incorporated the proposed method with the momentum training algorithm. Computer simulation results show that the proposed learning algorithm outperforms the conventional methods such as momentum and delta-bar-delta algorithms.

• Proceedings of the KIEE Conference
• /
• 1992.07a
• /
• pp.242-244
• /
• 1992

Conventionally in the industrial control, PlD controller has been used because of its robustness, and nonlinear characteristic of a system under control. Although the PlD controller produce suitable parameter of the each system and also variable of PlD controller should be changed according to environment, disturbance, load. In this paper, the convergence and learning accuracy of the back-propagation(BP) method in neural network are investigated by analyzing the reason for decelerating the convergence of BP method. and examining the rapid deceleration of the convergence when the learning is executed on the part of sigmoid activation function with the very small first derivative. The modified logistic activation function it proposed by defining the convergence factor based on the analysis and applied to the position and speed control of a DC-servo motor. This paper revealed for experimental, a neural network and a PD controller combined off-line system using developed the position and speed characteristics of a DC-servo motor.

• Journal of Broadcast Engineering
• /
• v.24 no.2
• /
• pp.234-242
• /
• 2019

In this paper, we propose a deep neural network which extracts SIFT feature points by determining whether the center pixel of a cropped image is a SIFT feature point. The data set of this network consists of a DIV2K dataset cut into $33{\times}33$ size and uses RGB image unlike SIFT which uses black and white image. The ground truth consists of the RobHess SIFT features extracted by setting the octave (scale) to 0, the sigma to 1.6, and the intervals to 3. Based on the VGG-16, we construct an increasingly deep network of 13 to 23 and 33 convolution layers, and experiment with changing the method of increasing the image scale. The result of using the sigmoid function as the activation function of the output layer is compared with the result using the softmax function. Experimental results show that the proposed network not only has more than 99% extraction accuracy but also has high extraction repeatability for distorted images.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2019.05a
• /
• pp.363-363
• /
• 2019

활성화 함수(activation function)는 기계학습(machine learning)의 학습과정에 비선형성을 도입하여 심층적인 학습을 용이하게 하고 예측의 정확도를 높이는 중요한 요소 중 하나이다(Roy et al., 2019). 일반적으로 기계학습에서 사용되고 있는 활성화 함수의 종류에는 계단 함수(step function), 시그모이드 함수(sigmoid 함수), 쌍곡 탄젠트 함수(hyperbolic tangent function), ReLU 함수(Rectified Linear Unit function) 등이 있으며, 예측의 정확도 향상을 위하여 다양한 형태의 활성화 함수가 제시되고 있다. 본 연구에서는 기계학습을 통하여 수위예측 시 정확도 향상을 위하여 Hybrid 활성화 함수를 제안하였다. 연구대상지는 조수간만의 영향을 받는 한강을 대상으로 선정하였으며, 2009년 ~ 2018년까지 10년간의 수문자료를 활용하였다. 수위예측 알고리즘은 Python 내 Tensorflow의 RNN (Recurrent Neural Networks) 모델을 이용하였으며, 강수량, 수위, 조위, 댐 방류량, 하천 유량의 수문자료를 학습시켜 3시간 및 6시간 후의 수위를 예측하였다. 예측정확도 향상을 위하여 입력 데이터는 정규화(Normalization)를 시켰으며, 민감도 분석을 통하여 신경망모델의 은닉층 개수, 학습률의 최적 값을 도출하였다. Hybrid 활성화 함수는 쌍곡 탄젠트 함수와 ReLU 함수를 혼합한 형태로 각각의 가중치($w_1,w_2,w_1+w_2=1$)를 변경하여 정확도를 평가하였다. 그 결과 가중치의 비($w_1/w_2$)에 따라서 예측 결과의 RMSE(Roote Mean Square Error)가 최소가 되고 NSE (Nash-Sutcliffe model Efficiency coefficient)가 최대가 되는 지점과 Peak 수위의 예측정확도가 최대가 되는 지점을 확인할 수 있었다. 본 연구는 현재 Data modeling을 통한 수위예측의 정확도 향상을 위해 기초가 되는 연구이나, 향후 다양한 형태의 활성화 함수를 제안하여 정확도를 향상시킨다면 예측 결과를 통하여 침수예보에 대한 의사결정이 가능할 것으로 기대된다.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.22 no.5
• /
• pp.1026-1035
• /
• 1997

In this paper, we studied for phoneme recognition using GPFN and PNN as a kind of RBFN. The structure of RBFN is similar to a feedforward networks but different from choosing of activation function, reference vector and learnign algorithm in a hidden layer. Expecially sigmoid function in PNN is replaced by one category included exponential function. And total calculation performance is high, because PNN performs pattern classification with out learning. In phonemerecognition experiment with 5 vowel and 12 consant, recognition rates of GPFN and PNN as a kind of RBFN reflected statistic characteristic of speech are higher than ones of MLP in case of using test data and quantizied data by VQ and LVQ.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.26 no.1
• /
• pp.87-92
• /
• 2016

In this paper, optimization technique such as particle swarm optimization was used to optimize the parameters of fuzzy Extreme Learning Machine. While the learning speed of conventional neural networks is very slow, that of Extreme Learning Machine is very fast. Fuzzy Extreme Learning Machine is composed of the Extreme Learning Machine with very fast learning speed and fuzzy logic which can represent the linguistic information of the field experts. The general sigmoid function is used for the activation function of Extreme Learning Machine. However, the activation function of Fuzzy Extreme Learning Machine is the membership function which is defined in the procedure of fuzzy C-Means clustering algorithm. We optimize the parameters of the membership functions by using optimization technique such as Particle Swarm Optimization. In order to validate the classification capability of the proposed classifier, we make several experiments with the various machine learning datas.

• Elastomers and Composites
• /
• v.50 no.2
• /
• pp.92-97
• /
• 2015

A study on reaction kinetics for a PTMG/TDI prepolymer with 2,2'-dichloro-4,4'-methylenedianiline (MOCA), of which formulations may be generally used for fabricating high performance polyurethane elastomers, was peformed using non-isothermal differential scanning calorimetry (DSC). A number of thermograms were obtained at several constant heating rates, and analysed using Flynn-Wall-Ozawa (FWO) isoconversional method for activation energy, $E_a$ and extended-Avrami equation for reaction order, n. Urea formation reaction of the present system was observed to occur through the simple exothermic reaction process in the temperature range of $100{\sim}130^{\circ}C$ for the heating rate of $3{\sim}7^{\circ}C/min$. and could be well-fitted with generalized sigmoid function. Though activation energy was nearly constant as $53.0{\pm}0.5kJ/mol$, it tended to increase a little at initial stage, but it decreases at later stage by the transformation into diffusion-controlled reaction due to the increased viscosity. Reaction order was evaluated as about 2.8, which was somewhat higher than the generally well-known $2^{nd}$ order values for the various urea reactions. Both the reaction order and reaction rate explicitly increased with temperature, which was considered as the indication of occurring the side reactions such as allophanate or biuret formation.

• Proceedings of the KIEE Conference
• /
• 1995.07b
• /
• pp.606-609
• /
• 1995

This paper presents discriminate between magnetizing inrush and internal faults of power transformer by artificial neural networks trained with preprocessing of fault discriminant. The proposed neural networks contain multi-layer perceptron using back-propagation learning algorithm with logistic sigmoid activation function. For this training and test, we used the relaying signals obtained from the EMTP simulation of model power system. It is shown that the proposed transformer protection system by neural networks never misoperated.