• 대한전기학회:학술대회논문집
• /
• 대한전기학회 2001년도 하계학술대회 논문집 D
• /
• pp.2030-2032
• /
• 2001

In recent years, many attempts have been made to predict the behavior of bonds, currencies, stock, or other economic markets. Most previous experiments used multilayer perceptrons(MLP) for stock market forecasting. The Kospi 200 Index is modeled using different neural networks and fuzzy system predictions. In this paper, a multilayer perceptron architecture, a dynamic polynomial neural network(DPNN) and a fuzzy system are used to predict the Kospi 200 index. The results of prediction is compared with the root mean squared error(RMSE) and the scatter plot. Results show that both networks can be trained to predict the index. And the fuzzy system is performing slightly better than DPNN and MLP.

• 대한전자공학회논문지
• /
• 제27권10호
• /
• pp.68-77
• /
• 1990

본 논문에서는 CAD 시스템의 신경망을 이용한 자동 입력기 구축을 위한 논리 심볼 인식방법을 제시한다. 손으로 작성한 도면을 인식하기 위해 특징 추출과 log mapping, 그리고 패턴 인식의 다단계 과정을 거친다. 각 논리 심볼의 현태 정보를 추출하기 위해 억제 가중치를 학습할 수 있는 경쟁 학습법을 제안하고 회전과 크기의 변화를 병진된 결과로 나타내는 log mapping을 하고 형태가 변한 심볼을 인식할 수 있도록 겹쳐지는 수용야(Receptive field)를 준비하여 error back propagation을 이용한 다층망으로 심볼을 인식한다.

• Nuclear Engineering and Technology
• /
• 제53권7호
• /
• pp.2334-2340
• /
• 2021

Radioactive particle tracking (RPT) is a minimally invasive nuclear technique that tracks a radioactive particle inside a volume of interest by means of a mathematical location algorithm. During the past decades, many algorithms have been developed including ones based on artificial intelligence techniques. In this study, RPT technique is applied in a simulated test section that employs a simplified mixer filled with concrete, six scintillator detectors and a¹³⁷Cs radioactive particle emitting gamma rays of 662 keV. The test section was developed using MCNPX code, which is a mathematical code based on Monte Carlo simulation, and 3516 different radioactive particle positions (x,y,z) were simulated. Novelty of this paper is the use of a location algorithm based on a deep learning model, more specifically a 6-layers deep rectifier neural network (DRNN), in which hyperparameters were defined using a Bayesian optimization method. DRNN is a type of deep feedforward neural network that substitutes the usual sigmoid based activation functions, traditionally used in vanilla Multilayer Perceptron Networks, for rectified activation functions. Results show the great accuracy of the DRNN in a RPT tracking system. Root mean squared error for x, y and coordinates of the radioactive particle is, respectively, 0.03064, 0.02523 and 0.07653.

• Computers and Concrete
• /
• 제10권6호
• /
• pp.557-573
• /
• 2012

This paper discusses the combined application of two different techniques, Neural Networks (NN) and Principal Component Analysis (PCA), for improved prediction of concrete properties. The combination of these approaches allowed the development of six neural networks models for predicting slump and compressive strength of concrete with mineral additives such as blast furnace slag, fly ash and silica fume. The Back-Propagation Multi-Layer Perceptron (BPMLP) with Bayesian regularization was used in all these models. They are produced to implement the complex nonlinear relationship between the inputs and the output of the network. They are also established through the incorporation of a huge experimental database on concrete organized in the form Mix-Property. Thus, the data comprising the concrete mixtures are much correlated to each others. The PCA is proposed for the compression and the elimination of the correlation between these data. After applying the PCA, the uncorrelated data were used to train the six models. The predictive results of these models were compared with the actual experimental trials. The results showed that the elimination of the correlation between the input parameters using PCA improved the predictive generalisation performance models with smaller architectures and dimensionality reduction. This study showed also that using the developed models for numerical investigations on the parameters affecting the properties of concrete is promising.

• Smart Structures and Systems
• /
• 제14권2호
• /
• pp.159-189
• /
• 2014

One of the most important requirements in the evaluation of existing structural systems and ensuring a safe performance during their service life is damage assessment. Damage can be defined as a weakening of the structure that adversely affects its current or future performance which may cause undesirable displacements, stresses or vibrations to the structure. The mass and stiffness of a structure will change due to the damage, which in turn changes the measured dynamic response of the system. Damage detection can increase safety, reduce maintenance costs and increase serviceability of the structures. Artificial Neural Networks (ANNs) are simplified models of the human brain and evolved as one of the most useful mathematical concepts used in almost all branches of science and engineering. ANNs have been applied increasingly due to its powerful computational and excellent pattern recognition ability for detecting damage in structural engineering. This paper presents and reviews the technical literature for past two decades on structural damage detection using ANNs with modal parameters such as natural frequencies and mode shapes as inputs.

• 스마트미디어저널
• /
• 제11권1호
• /
• pp.58-66
• /
• 2022

인터넷 기술이 발전함에 따라 SNS 사용자가 늘어나고 있다. SNS의 대중화가 진행되면서 소셜 네트워크의 영향력과 익명성을 활용한 SNS형 범죄가 나날이 증가하고 있는 추세이다. 본 논문에서는 인스타그램에서 SNS형 범죄에 주로 이용되는 가짜 계정 분류를 위해 통계 데이터와 이미지 데이터를 이용하여 각각 기계학습 및 딥러닝(deep learning) 기법을 활용한 가짜 계정 분류 방법을 제안한다. 모델 학습에 사용된 SNS 계정 데이터는 자체적으로 수집하였으며, 수집된 데이터는 통계 데이터 및 이미지 데이터에 기반한다. 통계 데이터의 경우에는 기계학습 및 다층 퍼셉트론 기반으로 학습을 진행하였고, 이미지 데이터의 경우에는 합성곱 신경망(Convolutional Neural Network, CNN) 기반으로 학습을 진행하였다. 학습을 진행한 결과 계정 분류에 대하여 정확도가 전반적으로 높게 나온 것을 확인하였다.

• Journal of Radiation Protection and Research
• /
• 제40권1호
• /
• pp.55-64
• /
• 2015

방사선치료계획장치의 핵심기술인 선량분포 계산은 빠르고 정확함을 요구한다. 기존 상용화된 치료계획장치의 선량 계산 방법은 빠르지만 정확성이 부족하고, 몬테칼로 방법은 시뮬레이션 시간과다 문제가 있다. 관심영역의 일부만 몬테칼로 방법이 계산하고 나머지 영역은 비선형함수사상 능력이 뛰어난 신경회로망이 계산하는 시스템은 상대적으로 빠르고 정확한 선량분포를 계산해낼 수 있다. 비균질 매질의 선량분포에 나타나는 불연속점과 변곡점의 특성을 신경회로망이 학습가능 하다는 것을 사전 작업을 통해 확인하였다. 이때 사용된 신경회로망은 Feedforward Multi-Layer Perceptron에 Scaled Conjugated Gradient 알고리즘과 Levenberg-Marquardt 알고리즘으로 각각 학습하여 성능비교를 하였고, 은닉층의 뉴런 개수에 따른 성능비교도 하였다. 마지막으로 균질매질의 팬텀에 대해 상용 치료계획장치의 선량계산 알고리즘으로 계산한 선량분포를 사전작업을 통해 확인된 신경회로망에 학습하여 깊이선량율의 평균제곱오차가 0.00214인 결과를 보여주었다. 균질 및 비균질 매질의 팬텀에 대한 3차원 선량분포를 계산하는 신경회로망 모델 개발 연구가 추가로 진행될 것이다.

• Earthquakes and Structures
• /
• 제5권4호
• /
• pp.417-438
• /
• 2013

This study was concerned on the application of a hybrid approach for analyzing the buried pipelines deformations subjected to earthquakes. Nonlinear time-history analysis of Finite Element (FE) model of buried pipelines, which was modeled using laboratory data, has been performed via selected earthquakes. In order to verify the FE model with experiments, a statistical test was done which demonstrated a good conformity. Then, the FE model was developed and the optimum intersection angle of pipeline and fault was obtained via genetic algorithm. Transient seismic strain of buried pipeline in the optimum intersection angle of pipeline and fault was investigated considering the pipes diameter, the distance of pipes from fault, the soil friction angles and seismic response duration of buried pipelines. Also, a two-layer perceptron Artificial Neural Network (ANN) was trained using results of FE model, and a nonlinear relationship was obtained to predict the bending strain of buried pipelines based on the pipes diameter, intersection angles of the pipelines and fault, the soil friction angles, distance of pipes from the fault, and seismic response duration; whereas it contains a wide range of initial input data without any requirement to laboratory measurements.

• 제어로봇시스템학회:학술대회논문집
• /
• 제어로봇시스템학회 2004년도 ICCAS
• /
• pp.652-657
• /
• 2004

As the real-time multimedia applications through Internet increase, the bandwidth available to TCP connections is oppressed by the UDP traffic, result in the performance of overall system is extremely deteriorated. Therefore, developing a new transmission protocol is necessary. The TCP-friendly algorithm is an example meeting this necessity. The TCP-friendly (TFRC) is an UDP-based protocol that controls the transmission rate based on the available round transmission time (RTT) and the packet loss rate (PLR). In the data transmission processing, transmission rate is determined based on the conditions of the previous transmission period. If the one-step ahead predicted values of the control factors are available, the performance will be improved significantly. This paper proposes a prediction model of transmission rate control factors that will be used for the transmission rate control, which improves the performance of the networks. The model developed through this research is predicting one-step ahead variables of RTT and PLR. A multiplayer perceptron neural network is used as the prediction model and Levenberg-Marquardt algorithm is used for the training. The values of RTT and PLR were collected using TFRC protocol in the real system. The obtained prediction model is validated using new data set and the results show that the obtained model predicts the factors accurately.

• 제어로봇시스템학회논문지
• /
• 제14권8호
• /
• pp.818-823
• /
• 2008

In this paper, we propose a feature extraction method which extracts directional features of handwritten numerals by using the projection runlength. Our directional featrures are obtained from four directional images, each of which contains horizontal, vertical, right-diagonal and left-diagonal lines in entire numeral shape respectively. A conventional method which extracts directional features by using Kirsch masks generates edge-shaped double line directional images for four directions, whereas our method uses the projections and their runlengths for four directions to produces single line directional images for four directions. To obtain the directional projections for four directions from a numeral image, some preprocessing steps such as thinning and dilation are required, but the shapes of resultant directional lines are more similar to the numeral lines of input numerals. Four [$4{\times}4$] directional features of a numeral are obtained from four directional line images through a zoning method. By using a hybrid feature which is made by combining our feature with the conventional features of a mesh features, a kirsch directional feature and a concavity feature, higher recognition rates of the handwrittern numerals can be obtained. For recognition test with given features, we use a multi-layer perceptron neural network classifier which is trained with the back propagation algorithm. Through the experiments with the handwritten numeral database of Concordia University, we have achieved a recognition rate of 97.85%.