• 한국환경복원기술학회지
• /
• 제10권6호
• /
• pp.70-80
• /
• 2007

This study was surveyed from January to september, 2007 in order to evaluate the function of coastal wetland as a ecological axis in korea peninsula. Assessment was done by RAM (Rapid Assessment Method). RAM is consisted of total 8 contents and divided into high, moderate, low. The preservation grade of RAM is divided into 4 grades; absolute preservation, preservation, improvement and improvement or restoration. Throughout survey on total 14 wetlands of marine, estuary wetland and back marsh which are distributed in west coast in chung-nam province, their function was assessed. As result, total all the 14 wetlands were judged as preservation grade by assessment of 8 functional contents. The function of wetlands assessed as preservation grade showed high in water quality protection and improvement. Also, showed high in vegetation diversity, wildlife habitat and aesthetic recreation. Meanwhile, showed low in ground water recharge, Shoreline/Stream Bank Protection, Flood/Stormwater storage and Flood flow alteration. Of wetlands evaluated as preservation grade, Dae-ho, Sinduri, Bu-Nam lake, Sowhang dune and keum river estuary were assessed as absolute preservative area owing to habitation of international protection species and endangered species. These wetlands are essential to be managed continuously as a area having high ecological value. Farther, this wetlands will be done as a axis of ecological network related to land ecosystem.

• Journal of Electrical Engineering and Technology
• /
• 제13권6호
• /
• pp.2468-2478
• /
• 2018

We propose the Sparse Feature Convolutional Neural Network (SFCNN) to reduce the volume of convolutional neural networks (CNNs). Despite the superior classification performance of CNNs, their enormous network volume requires high computational cost and long processing time, making real-time applications such as online-training difficult. We propose an advanced network that reduces the volume of conventional CNNs by producing a region-based sparse feature map. To produce the sparse feature map, two complementary region-based value extraction methods, cluster max extraction and local value extraction, are proposed. Cluster max is selected as the main function based on experimental results. To evaluate SFCNN, we conduct an experiment with two conventional CNNs. The network trains 59 times faster and tests 81 times faster than the VGG network, with a 1.2% loss of accuracy in multi-class classification using the Caltech101 dataset. In vehicle classification using the GTI Vehicle Image Database, the network trains 88 times faster and tests 94 times faster than the conventional CNNs, with a 0.1% loss of accuracy.

• 한국조경학회지
• /
• 제29권1호
• /
• pp.67-79
• /
• 2001

The purpose of this research is to evaluate the biotope value in a metropolitan area after constructing the biotope evaluation model to develop biotope networking in the light of ecological landscape planning. In addition this study is to provide a basic frame of constructing the biotope networking system in the future with the evaluation results. The scope of this research is limited to high density residential area including some part of the forest adjacent to urban boundary. The results of the classified biotope in the research area have been turned out to be 24 patterns. According to the result of estimated nature value having the forth grade, three were forests combined with urban natural parks, schools near forests, and grasslands. There is a little place having first grade. We an deduce this fact that we have surveyed biotopes, having the minimum nature value and wider than children's park. As for the result of evaluated the structure, there were the forests of Seong-Dong, Meaho-Dong, and Kumho river, which is the forth grade. As for the evaluation of connection-function, the followings area were evaluated as the third grade-linear biotopes having highly valued and sustainable similarity - Seoul-Pusan Railroad, Dong-Daegu St., Kumho river, river basins. Bum대 park and a botanic park near Suesung lake, As for the evaluated the buffer-function, higly valued production biotopes, existing at a transition belt between he residential density areas, are Sawol-Dong, Meaho-Dong and Kumho river. When each of the indexes was evaluated, the core of the biotop contained most of natural biotopes. Large artificial biotopes were evaluated as an important biotopes, while small artificial biotopes were classified as a dot biotope. The future research on the concrete biotope networking construction, based upon performed evaluation in this research, should be conducted.

• 대한산업공학회지
• /
• 제24권2호
• /
• pp.251-260
• /
• 1998

This paper deals with a simulated annealing neural network model for determining sequences of models inputted into a mixed model assembly line. We first present a energy function fitting to our problem, next determine the value of the parameters of the energy function using convergence ratio and the number of searched feasible solution. Finally we compare our model NMS with the modified Thomopoulos model. The result of the comparison shows that NMS and Thomopoulos offer a similar output in the problems having good smoothness.

• 한국신뢰성학회지:신뢰성응용연구
• /
• 제8권4호
• /
• pp.181-192
• /
• 2008

In this paper, we proposed effective algorithm that do Availability of communication network account using total probability formula. Use path tracing method about PSTN linked with MSC in WCDMA considering occasion that do not consider mobility of terminal and Availability did account. First, express minimum path series gathering by structure function after network that do modeling by stochastic graph for communication network saves all minimal path sets between particular two abutments and we gain Availability value in angular variable after express structure function by Boolean operation form and Availability between two abutments did account.

• International journal of advanced smart convergence
• /
• 제8권1호
• /
• pp.133-140
• /
• 2019

In this paper, we present an algorithm for adjusting degree of belief for consistency on the evidence dependency network where various sets of evidence support different sets of hypotheses. It is common for experts to assign higher degree of belief to a hypothesis when there is more evidence over the hypothesis. Human expert without knowledge of uncertainty handling may not be able to cope with how evidence is combined to produce the anticipated belief value. Belief in a hypothesis changes as a series of evidence is known to be true. In non-monotonic reasoning environments, the belief retraction method is needed to clearly deal with uncertain situations. We create evidence dependency network from rules and apply the evidence retraction algorithm to refine belief values on the hypothesis set. We also introduce negative belief values to reflect the reverse effect of evidence combination.

• Journal of Communications and Networks
• /
• 제12권6호
• /
• pp.600-604
• /
• 2010

A method of width adaptation in the radial basis function network (RBFN) using stochastic gradient (SG) algorithm is introduced. Using Taylor's expansion of error signal and differentiating the error with respect to the step-size, the optimal time-varying step-size of the width in RBFN is derived. The proposed approach to adjusting widths in RBFN achieves superior learning speed and the steady-state mean square error (MSE) performance in nonlinear channel environment. The proposed method has shown enhanced steady-state MSE performance by more than 3 dB in both nonlinear channel environments. The results confirm that controlling over step-size of the width in RBFN by the proposed algorithm can be an effective approach to enhancement of convergence speed and the steady-state value of MSE.

• 전자공학회논문지A
• /
• 제28A권10호
• /
• pp.782-790
• /
• 1991

In this paper, we implement the function of the IEEE802.2LLC(logical link control) sublayer one of the 7layers in MAP(manufacturing automation protocol), a standard communication protocol for manufacturing automation. With the assembly language we designed the class 3 function based on the IEEE standards and verified on the network adaptor hardware. In this experiment, we tested the function using the simplified variables considering that the retransmission value was chosen to be 1 and the life-time of the transmission variable infinite. According to the result of the service procedures, we confirmed that user data were transmitted to the corresponding station without any error.

• 디지털콘텐츠학회 논문지
• /
• 제6권1호
• /
• pp.7-12
• /
• 2005

논문은 IEEE 표준인 SCI(Scalable Coherent Interface)에 근거한 degree 2의 ICR(Interleaved Cyclic Ring) 네트워크의 신뢰도를 연구하였고 대표적인 링 네트워크들과 비교하였다. ICR 네트워크의 신뢰도 연구에 있어서 두 노드가 동시에 통신 할 수 있는 신뢰도를 계산하였다. ICR 네트워크의 신뢰도 연구에서 실패율(failure rate)의 변화에 신뢰도의 반응을 연구하였고 적은 노드에 대하여 정확한 신뢰도를 구하고 노드 수가 증가한 경우에는 최대 최소 경계를 계산하여 평균치 일반식을 구하였다. ICR 네트워크의 사이클의 변화에 따라 ICR 네트워크의 신뢰도를 다른 링 네트워크와 비교하였다.

• International journal of advanced smart convergence
• /
• 제13권2호
• /
• pp.80-87
• /
• 2024

This paper focuses on improving accuracy in constrained computing settings by employing the ReLU (Rectified Linear Unit) activation function. The research conducted involves modifying parameters of the ReLU function and comparing performance in terms of accuracy and computational time. This paper specifically focuses on optimizing ReLU in the context of a Multilayer Perceptron (MLP) by determining the ideal values for features such as the dimensions of the linear layers and the learning rate (Ir). In order to optimize performance, the paper experiments with adjusting parameters like the size dimensions of linear layers and Ir values to induce the best performance outcomes. The experimental results show that using ReLU alone yielded the highest accuracy of 96.7% when the dimension sizes were 30 - 10 and the Ir value was 1. When combining ReLU with the Adam optimizer, the optimal model configuration had dimension sizes of 60 - 40 - 10, and an Ir value of 0.001, which resulted in the highest accuracy of 97.07%.