• Proceedings of the Korean Institute of Intelligent Systems Conference
• /
• 1996.10a
• /
• pp.239-242
• /
• 1996

본 논문에서는 BP(Back Propagation)에 비해서 빠른 학습시간과 다른 경쟁학습 신경회로망 알고리즘에 비해서 비교적 우수한 성능으로 패턴인식 등에 많이 이용되고 있는 LVQ(Learning Vector Quantization) 알고리즘의 성능을 향상시키기 위한 방법을 논의하고자 한다. 일반적으로 LVQ는 음(negative)의 학습을 하기 때문에 초기 가중치가 제대로 설정되지 않으면 발산할 수 있다는 단점이 있으며, 경쟁학습 계열의 신경망이기 때문에 출력 층의 뉴런 수에 따라 성능에 큰 영향을 받는다고 알려져 있다.[1]. 지도학습 형태를 지닌 LVQ의 경우에 학습패턴이 n개의 클래스를 가지고, 각 클래스 별로 학습패턴의 수가 같은 경우에 일반적으로 전체 출력뉴런에 대해서 (출력뉴런수/n)개의 뉴런을 각 클래스의 목표(desired) 클러스터로 할당하여 학습을 수행하는데, 본 논문에서는 각 클래스에 동일한 수의 출력뉴런을 할당하지 않고, 학습데이터에서 각 클래스의 분산을 추정하여 각 클래스의 분산을 추정분산에 비례하게 목표 출력뉴런을 할당하고, 초기 가중치도 추정분산에 비례하게 각 클래스의 초기 임의 위치 입력백터를 사용하여 학습을 수행하는 방법을 제안한다. 본 논문에서 제안하는 방법은 분류하고자 하는 데이터에 대해서 필요한 최적의 출력뉴런 수를 찾는 것이 아니라 이미 결정되어 있는 출력뉴런 수에 대해서 각 클래스에 할당할 출력 뉴런 수를 데이터의 추정분산에 의해서 결정하는 것으로, 추정분산이 크면 상대적으로 많은 출력 뉴런을 할당하고 작으면 상대적으로 적은 출력뉴런을 할당하고 초기 가중치도 마찬가지 방법으로 결정하며, 이렇게 하면 정해진 출력뉴런 개수 안에서 각 클래스 별로 분류의 어려움에 따라서 출력뉴런을 할당하기 때문에 미학습 뉴런이 줄어들게 되어 성능의 향상을 기대할 수 있으며, 실험적으로 제안된 방법이 더 나은 성능을 보임을 확인했다.initially they expected a more practical program about planting than programs that teach community design. Many people are active in their own towns to create better environments and communities. The network system "Alpha Green-Net" is functional to support graduates of the course. In the future these educational programs for citizens will becomes very important. Other cities are starting to have their own progrms, but they are still very short term. "Alpha Green-Net" is in the process of growing. Many members are very keen to develop their own abilities. In the future these NPOs should become independent. To help these NPOs become independent and active the educational programs should consider and teach about how to do this more in the future.단하였는데 그 결과, 좌측 촉각엽에서 제4형의 신경연접이 퇴행성 변화를 나타내었다. 그러므로 촉각의 지각신경세포는 뇌의 같은 족 촉각엽에 뻗어와 제4형 신경연접을 형성한다고 결론되었다.$/ 값이 210 $\mu\textrm{g}$/$m\ell$로서 효과적인 저해 활성을 나타내었다 따라서, 본 연구에서 빈

• International Journal of Oral Biology
• /
• v.31 no.4
• /
• pp.113-118
• /
• 2006

In the primary sensory neuron of the mesencephalic trigeminal nucleus (MTN), the peripheral axon supplies a large number of annulospiral endings surrounding intrafusal fibers encapsulated in single muscle spindles while the central axon sends only a few number of synapses onto single ${\alpha}-motoneurons({\alpha}-MNs)$. Therefore, the ${\alpha}-{\gamma}$ linkage is thought to be very crucial in the jaw-closing movement. Spike activity in a ${\gamma}-motoneuron\;({\gamma}-MN)$ would induce a large number of impulses in single peripheral axons by activating many intrafusal fibers simultaneously, subsequently causing an activation of ${\alpha}-MNs$ in spite of the small number of synapses. Thus, the activity of ${\gamma}-MNs$ may be vital for modulation of jaw-closing movements. Independently of such a spindle activity modulated by ${\gamma}-MNs$, somatic depolarization in MTN neurons is known to trigger the oscillatory spike activity. Nevertheless, the trafficking of these spikes arising from the two distinct sources of MTN neurons is not well understood. In this short review, switching among multiple functional modes of MTN neurons is discussed. Subsequently, it will be discussed which mode can support the ${\alpha}-{\gamma}$ linkage. In our most recent study, simultaneous patch-clamp recordings from the soma and axon hillock revealed a spike-back-propagation from the spike-initiation site in the stem axon to the soma in response to a somatic current pulse. The persistent $Na^+$ current was found to be responsible for the spike-initiation in the stem axon, the activation threshold of which was lower than those of soma spikes. Somatic inputs or impulses arising from the sensory ending, whichever trigger spikes in the stem axon first, would be forwarded through the central axon to the target synapse. We also demonstrated that at hyperpolarized membrane potentials, 4-AP-sensitive $K^+$ current ($IK_{4-AP}$) exerts two opposing effects on spikes depending on their origins; the suppression of spike initiation by increasing the apparent electrotonic distance between the soma and the spike-initiation site, and the facilitation of axonal spike invasion at higher frequencies by decreasing the spike duration and the refractory period. Through this mechanism, the spindle activity caused by ${\gamma}-MNs$ would be safely forwarded to ${\alpha}-MNs$. Thus, soma spikes shaped differentially by this $IK_{4-AP}$ depending on their origins would reflect which one of the two inputs was forwarded to the target synapses.

• Journal of the Korean Geotechnical Society
• /
• v.21 no.8
• /
• pp.73-84
• /
• 2005

A three layered neural network model was developed using back propagation algorithm to estimate the UU undrained shear strength of Korean soft soil based on the database of actual undrained shear strengths and piezocone measurements compiled from 8 sites over the Korea. The developed model was validated by comparing model predictions with measured values about new piezocone data, which were not previously employed during development of model. Performance of the neural network model was also compared with conventional empirical methods. It was found that the number of neuron in hidden layer is different for the different combination of transfer functions of neural network models. However, all piezocone neural network models are successful in inferring a complex relationship between piezocone measurements and the undrained shear strength of Korean soft soils, which give relatively high coefficients of determination ranging from 0.69 to 0.72. Since neural network model has been generalized by self-learning from database of piezocone measurements and undrained shear strength over the various sites, the developed neural network models give more precise and generally reliable undrained shear strengths than empirical approaches which still need site specific calibration.

• The Transactions of the Korea Information Processing Society
• /
• v.4 no.2
• /
• pp.465-482
• /
• 1997

In this paper, we have studied on cinstructing code-recognition shstem by neural network according to a image process taking the DOT classification code stamped on tire surface.It happened to a few problems that characters distorted in edge by diffused reflection and two adjacent characters take the same label,even very sen- sitive to illumination ofr recognition the stamped them on tire.Thus,this paper would propose the algorithm for tire code under being cinscious of these properties and prove the algorithm drrciency with a simulation.Also,we have suggerted the MCBP network composing of multi-linked recognizers of dffcient identify the DOT code being tire classification code.The MCBP network extracts the projection balue for classifying each character's rdgion after taking out the prjection of each chracter's region on X,Y axis,processes each chracters by taking 7$\times$8 normalization.We have improved error rate 3% through the MCBP network and post-process comparing the DOT code Database. This approach has a accomplished that learming time get's improvenent at 60% and recognition rate has become to 95% from 90% than BckPropagation with including post- processing it has attained greate rates of entire of tire recoggnition at 98%.

• Geophysics and Geophysical Exploration
• /
• v.14 no.3
• /
• pp.220-226
• /
• 2011

In the elastic wave equations, both horizontal and vertical displacements are defined. Since we can measure both the horizontal and vertical displacements in field acquisition, these displacements compose a displacement vector. In this study, we propose a frequency-domain elastic waveform inversion technique taking advantage of the magnitudes of displacement vectors to define objective function. When we apply this displacement-vector objective function to the frequency-domain waveform inversion, the inversion process naturally incorporates the back-propagation algorithm. Through the inversion examples with the Marmousi model and the SEG/EAGE salt model, we could note that the RMS error of the solution obtained by our algorithm decreased more stably than that of the conventional method. Particularly, the density of the Marmousi model and the low-velocity sub-salt zone of the SEG/EAGE salt model were successfully recovered. Since the gradient direction obtained from the proposed objective function is numerically unstable, we need additional study to stabilize the gradient direction. In order to perform the waveform inversion using the displacementvector objective function, it is necessary to acquire multi-component data. Hence, more rigorous study should be continued for the multi-component land acquisition or OBC (Ocean Bottom Cable) multi-component survey.

• Journal of Biomedical Engineering Research
• /
• v.19 no.5
• /
• pp.469-476
• /
• 1998

In the case of acquisition and displaying medical Images such as ultrasonography and endoscopy on VGA monitor of PC system, image degradation of tear-drop appears through scan conversion. In this study, we compare several methods which can solve this degradation and implement the hardware system that resolves this problem in real-time with PC. It is possible to represent high quality image display and real-time processing and acquisition with specific de-interlacing device and PCI bridge on our hardware system. Image quality is improved remarkably on our hardware system. It is implemented as PC-based system, so acquiring, saving images and describing text comment on those images and PACS networking can be easily implemented.metabolism. All images were spatially normalized to MNI standard PET template and smoothed with 16mm FWHM Gaussian kernel using SPM96. Mean count in cerebral region was normalized. The VOls for 34 cerebral regions were previously defined on the standard template and 17 different counts of mirrored regions to hemispheric midline were extracted from spatially normalized images. A three-layer feed-forward error back-propagation neural network classifier with 7 input nodes and 3 output nodes was used. The network was trained to interpret metabolic patterns and produce identical diagnoses with those of expert viewers. The performance of the neural network was optimized by testing with 5~40 nodes in hidden layer. Randomly selected 40 images from each group were used to train the network and the remainders were used to test the learned network. The optimized neural network gave a maximum agreement rate of 80.3% with expert viewers. It used 20 hidden nodes and was trained for 1508 epochs. Also, neural network gave agreement rates of 75~80% with 10 or 30 nodes in hidden layer. We conclude that artificial neural network performed as well as human experts and could be potentially useful as clinical decision support tool for the localization of epileptogenic zones.

• Journal of Intelligence and Information Systems
• /
• v.8 no.2
• /
• pp.205-227
• /
• 2002

Our study is focused on a multiple-agent system to provide efficient collaborative work by automating the conference calling process with the help of intelligent agents. Automating the meeting scheduling requires a careful consideration of the individual official schedule as well as the privacy and personal preferences. Therefore, the automation of conference calling needs the distributed processing task where a separate calendar management process is associated for increasing the reliability and inherent parallelism. This paper describes in detail the design and implementation issues of a multiple-agent system for conference calling that allows the convener and participants to minimize their efforts in creating a meeting. Our system is based on the client-sewer model. In the sewer side, a scheduling agent, a negotiating agent, a personal information managing agent, a group information managing agent, a session managing agent, and a coordinating agent are operating. In the client side, an interface agent, a media agent, and a collaborating agent are operating. Agents use a standardized knowledge manipulation language to communicate amongst themselves. Communicating through a standardized knowledge manipulation language allows the system to overcome heterogeneity which is one of the most important problems in communication among agents for distributed collaborative computing. The agents of our system propose the dates on which as many participants as possible are available to attend the conference using the forward chaining algorithm and the back propagation network algorithm.

• Resources Recycling
• /
• v.29 no.3
• /
• pp.11-23
• /
• 2020

This study aims to investigate breakage and liberation characteristics of iron ore from Shinyemi mine, Jeongseon by ball mill. Parameters of breakage functions for three grade samples of iron ore were obtained using single-sized-feed breakage test and back-calculation based on nonlinear programming. The results showed that with the increase in the grade of iron ore, the breakage rate factor decrease whereas the particle size sensitivity decreases. This results from retardation of microcrack-propagation by magnetite grain in the ore. Breakage distribution analysis showed that the breakage mechanism appear to be impact fracture dominant with the increase of grade owing to the stress distribution effect by magnetite grain. Degree of liberation (DOL) increased with the increase in grade and decrease in particle size, respectively. Using the breakage function and size-DOL relationship, a model that can predict time-dependent-DOL is established. When scale-up factors from operating condition are available, the model is expected to be capable of predicting size and DOL with time in actual mining process.

• The Transactions of the Korea Information Processing Society
• /
• v.4 no.7
• /
• pp.1749-1758
• /
• 1997

This paper suggests the development of dynamic forecasting model for short-term power demand based on Radial Basis Function Network and Pal's GLVQ algorithm. Radial Basis Function methods are often compared with the backpropagation training, feed-forward network, which is the most widely used neural network paradigm. The Radial Basis Function Network is a single hidden layer feed-forward neural network. Each node of the hidden layer has a parameter vector called center. This center is determined by clustering algorithm. Theatments of classical approached to clustering methods include theories by Hartigan(K-means algorithm), Kohonen(Self Organized Feature Maps %3A SOFM and Learning Vector Quantization %3A LVQ model), Carpenter and Grossberg(ART-2 model). In this model, the first approach organizes the load pattern into two clusters by Pal's GLVQ clustering algorithm. The reason of using GLVQ algorithm in this model is that GLVQ algorithm can classify the patterns better than other algorithms. And the second approach forecasts hourly load patterns by radial basis function network which has been constructed two hidden nodes. These nodes are determined from the cluster centers of the GLVQ in first step. This model was applied to forecast the hourly loads on Mar. $4^{th},\;Jun.\;4^{th},\;Jul.\;4^{th},\;Sep.\;4^{th},\;Nov.\;4^{th},$ 1995, after having trained the data for the days from Mar. $1^{th}\;to\;3^{th},\;from\;Jun.\;1^{th}\;to\;3^{th},\;from\;Jul.\;1^{th}\;to\;3^{th},\;from\;Sep.\;1^{th}\;to\;3^{th},\;and\;from\;Nov.\;1^{th}\;to\;3^{th},$ 1995, respectively. In the experiments, the average absolute errors of one-hour ahead forecasts on utility actual data are shown to be 1.3795%.

• Annual Conference of KIPS
• /
• 2001.10a
• /
• pp.727-730
• /
• 2001

멀티미디어 환경의 급속한 발전에 의해 영상처리 기술은 인간의 인체와 관련하여 얼굴인식, 제스처 인식에 관한 응용과 더불어 스포츠 관련분야로 깊숙히 정착하고 있다. 그러나 입력영상으로부터 움직이고 있는 선수들의 동작을 추출 및 추적하는 일은 컴퓨터비전 연구의 난 문제 중의 하나로 알려져 있다. 이러한 축구경기의 TV 중계에 있어서 하이라이트 장면의 자동추출(자동색인)은 그 경기의 가장 집약적인 표현이며, 축구경기 전체를 한 눈에 파악할 수 있도록 해주는 요약(summary)이자 intensive actions이고 경기의 진수이다. 따라서 축구경기와 같이 비교적 기 시간(대체로 1시간 30분) 동안 다수의 선수(양 팀 합해서 22명)들이 서로 복잡하게 뒤얽히면서 진행하는 경기의 하이라이트 장면을 효과적으로 포착하여 표현해 줄 수 있다면 TV를 통해서 경기를 관람하는 시청자들에게는 경기의 진행상황을 한 눈에 효과적으로 파악할 수 있게 해주어 흥미진진한 경기관람을 할 수 있게 해주고, 경기의 진행자들(감독, 코치, 선수 등)에게는 고차원적이고 과학적인 정보를 효과적으로 제공함으로써 한층 진보된 경기기법을 개발하고 과학적인 경기전략을 세울 수 있게 해준다. 본 논문은 이상과 같이 팀 스포츠(Team Spots)의 일종인 축구경기 하이라이트 장면의 자동색인을 위해 뉴럴네트워크 기법을 이용하여 그룹 포메이션(Group Formation) 중의 공격패턴 자동분류 기법을 개발하고 이를 검증하였다. 본 연구에서는 축구경기장 내의 빈번하게 변화하는 장면들을 자동으로 분할하여 대표 프레임을 선정하고, 대표 프레임 상에서 선수들의 위치정보와 공의 위치정보 등을 기초로 하여 경기 중에 이루어지는 선수들의 그룹 포메이션을 추적하여 그룹행동(group behavior)을 분석하고, 뉴럴네트워크의 BP(Back-Propagation) 알고리즘을 사용하여 축구경기 공격패턴을 자동으로 인식 및 분류함으로써 축구경기 하이라이트 장면의 자동추출을 위한 기반을 마련하였다. 본 연구의 실험에는 '98 프랑스 월드컵 축구경기의 다양한 공격패턴에 대한 비디오 영상에서 각각 좌측공격 60개, 우측공격 74개, 중앙공격 72개, 코너킥 39개, 프리킥 52개의 총 297개의 데이터를 추출하여 사용하였다. 실험과는 좌측공격 91.7%, 우측공격 100%, 중앙공격 87.5%, 코너킥 97.4%, 프리킥 75%로서 매우 양호한 인식율을 보였다.