• Spatial Information Research
• /
• v.2 no.2
• /
• pp.175-188
• /
• 1994

A competitive learning network was proposed as unsupervised training method of remote sensing data, Its performance and computational re¬quirements were compared with conventional clustering techniques such as Se¬quential and K - Means. An airborne remote sensing data set was used to study the performance of these classifiers. The proposed algorithm required a little more computational time than the conventional techniques. However, the perform¬ance of competitive learning network algorithm was found to be slightly more than those of Sequential and K - Means clustering techniques.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.10 no.3
• /
• pp.203-210
• /
• 2000

In this paper, we propose a new unsupervised learning network and competitive learning algorithm for pattern classification. The proposed network is based on relative similarity, which is similarity measure between input data and cluster group. So, the proposed network and algorithm is called relative similarity network(RSN) and learning algorithm. According to definition of similarity and learning rule, structure of RSN is designed and pseudo code of the algorithm is described. In general pattern classification, RSN, in spite of deletion of learning rate, resulted in the identical performance with those of WTA, and SOM. While, in the patterns with cluster groups of unclear boundary, or patterns with different density and various size of cluster groups, RSN produced more effective classification than those of other networks.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2001.10a
• /
• pp.581-584
• /
• 2001

신경망이 만들어내는 출력에 대한 정보는 수치적으로 분산되어 신경망에 저장되므로, 인간이 직접 해석하기가 힘들다. 본 논문에서는 LRE(link rule extraction)기법인 NofM 알고리즘의 6단계 중에서 초기 단계인 가중치 군집화 단계를 개선하여 추출되는 규칙들의 전제부에 들어가는 규칙 조건들의 수를 조절함으로써, 추출된 규칙이 입력 특성에 대한 정보를 과잉 일반화하거나, 과잉 구체화하는 것을 피할 수 있음을 실험을 통해 보였다. 일반적으로 NofM 알고리즘에서 가중치들을 군집화한 때는 Join 알고리즘을 사용하는데, 본 논문에서는 Join 알고리즘의 Join condition을 0.05부터 0.25까지 0.05씩 점진적으로 확대하여 클러스터링을 하여줌으로써 신경망의 출력에 중요한 역할을 하는 가중치들을 효과적으로 군집화함을 보였다.

• KSCE Journal of Civil and Environmental Engineering Research
• /
• v.29 no.2B
• /
• pp.111-120
• /
• 2009

A local scour around a bridge pier is known as one of important factors of bridge collapse. Two approaches are usually used in estimating a scour depth in practice. One is to use empirical formulas, and the other is to use computational methods. But the use of empirical formulas is limited to predict a scour depth under similar conditions to which the formulas were derived. Computational methods are currently too expensive to be applied to practical engineering problems. This study presented the application of artificial neural networks (ANN) to the prediction of a scour depth around a bridge pier at an equilibrium state. This study also investigated various ANN algorithms for estimating a scour depth, such as Backpropagation Network, Radial Basis Function Network, and Generalized Regression Network. Preliminary study showed that ANN models resulted in very wide range of errors in predicting a scour depth. To solve this problem this study incorporated cluster analysis into ANN. The incorporation of cluster analysis provided better estimations of scour depth up to 42% compared with other approaches.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2002.04a
• /
• pp.519-522
• /
• 2002

일반적으로 PubMed와 같은 인터넷을 이용한 대규모 의료 문헌정보 검색시스템에서 포괄적인 주제어나 간결한 주제어를 이용한 검색을 시도할 경우, 종종 매우 다양한 세부주제의 문헌리스트들이 다량으로 검색된다. 이러한 경우 이용자는 실제로 본인이 원했던 세부주제에 부합되는 문헌들을 찾기 위해서는 검색결과로 주어진 긴 문헌리스트상의 문헌 하나하나에 대해 다시 문헌제목이나 혹은 요약 등의 내용을 직접 읽어보고 내용을 확인하여야 한다. 이러한 작업은 매우 번거럽고 시간과 노력을 많이 필요로 한다. 따라서 본 논문에서는 이러한 노력을 줄이기 위한 한 가지 방안으로, PubMed 시스템의 주제어 검색결과로 주어진 문헌들에 대해 내용의 유사성과 차별성에 따라 자동으로 몇 개의 그룹으로 나누어주는 군집화시스템 MedCluster의 설계와 구현에 대해 소개한다. MedCluster의 큰 특징은 기존의 문서 군집화 방법과는 다른 신경망 GHSOM을 이용한 군집화 방법을 사용하는 점이다. GHSOM은 미리 문서 그룹의 개수를 정해줄 필요가 없고 다양한 레벨의 문서 그룹들을 얻을 수 있는 계층적 군집화를 이루어낸다는 장점을 가지고 있다. 본 논문에서는 신경망 GHSOM의 구조와 특성에 대해 간략히 살펴보고, GHSOM을 채용한 의료문헌 군집화시스템 MedCluster의 설계와 구현에 대해 설명한다.

• KSCE Journal of Civil and Environmental Engineering Research
• /
• v.29 no.3C
• /
• pp.115-125
• /
• 2009

The preconsolidation pressure has been commonly determined by oedometer test. However, it can also be determined by insitu test, such as piezocone test with theoretical and(or) empirical correlations. Recently, Neural Network (NN) theory was applied and some models were proposed to estimate the preconsolidation pressure or OCR. It was already found that NN model can come over the site dependency and prediction accuracy is greatly improved when compared with present theoretical and empirical models. However, since the optimization process of synaptic weights of NN model is dependent on the initial synaptic weights, NN models which are trained with different initial weights can't avoid the variability on prediction result for new database even though they have same structure and use same transfer function. In this study, Committee Neural Network (CNN) model is proposed to improve the initial weight dependency of multi-layered neural network model on the prediction of preconsolidation pressure of soft clay from piezocone test result. Prediction results of CNN model are compared with those of conventional empirical and theoretical models and multi-layered neural network model, which has the optimized structure. It was found that even though the NN model has the optimized structure for given training data set, it still has the initial weight dependency, while the proposed CNN model can improve the initial weight dependency of the NN model and provide a consistent and precise inference result than existing NN models.

• The Journal of Information Technology
• /
• v.5 no.2
• /
• pp.17-26
• /
• 2002

On this paper, we propose SOSL(Self-Organized Supervised Learning) and it's architecture SOSL is hybrid type neural network. It consists of several CBP (Component Back Propagation) neural networks, and a modified PCA neural networks. CBP neural networks perform supervised learning procedure in parallel to clustered and complex input patterns. Modified PCA networks perform it's learning in order to transform dimensions of original input patterns to lower dimensions by clustering and local projection. Proposed SOSL can effectively apply to neural network learning with large input patterns results in huge networks size.

• Proceedings of the Korea Inteligent Information System Society Conference
• /
• 2000.11a
• /
• pp.479-488
• /
• 2000

다양한 재무정보를 이용하여 기업간 경쟁적 벤치마킹을 수행하는 것은 매우 어려운 작업인 동시에 분석에 상담한 시간이 소요된다. 본 연구에서는 재무정보를 이용한 기업간 경쟁적 벤치마킹을 효과적으로 수행하기 위하여 대표적인 자율신경망 모형인 자기조직화 신경망을 분석에 이용하였다. 자기조직화 신경망은 다차원적인 재무자료를 2차원 출력 공간으로 투영함으로써 결과를 시각화하는데 매우 효과적이며, 시각화된 결과는 재무적인 경쟁우위에 따라 기업을 군집화함으로써 효과적인 경쟁적 벤치마킹을 수행할 수 있도록 한다. 본 연구에서는 1998년. 1999년, 그리고 2000년 상반기까지의 국내 제조업체 재무구조 분석사례에 자기조직화 신경함을 적용하여 재무적 경쟁우위에 따른 기업들의 군집화 모형으로서의 가능성을 제시하였다.

• Proceedings of the Korean Information Science Society Conference
• /
• 2000.04b
• /
• pp.475-477
• /
• 2000

본 논문에서는 적응적 성분분석 기법을 이용하여 radial basis 함수 신경망의 학습시간과 분류성능을 개선한 새로운 기법을 제안하였다. 제안된 기법에서 적응적 성분분석 기법은 radial basis 함수 신경망의 은닉층 뉴런 개수와 중심값 설정을 위해 이용하였다. 제안된 기법의 radial basis 함수 신경망을 200명의 암환자를 2부류(초기와 악성)로 분류하는 문제에 적용하여 시뮬레이션한 결고, k-평균 군집화 알고리즘을 이용한 radial basis 함수 신경망과 비교할 때 학습시간과 시험 데이터의 분류에서 더욱 우수한 성능이 있음을 확인할 수 있었다.

• Investigative Magnetic Resonance Imaging
• /
• v.6 no.2
• /
• pp.137-146
• /
• 2002

Purpose : This paper introduces an improved classification approach which adopts a deterministic relaxation method and an agglomerative clustering technique for the classification of MRI using neural network. The proposed approach can solve the problems of convergency to local optima and computational burden caused by a large number of input patterns when a neural network is used for image classification. Materials and methods : Application of Hopfield neural network has been solving various optimization problems. However, major problem of mapping an image classification problem into a neural network is that network is opt to converge to local optima and its convergency toward the global solution with a standard stochastic relaxation spends much time. Therefore, to avoid local solutions and to achieve fast convergency toward a global optimization, we adopt MFA to a Hopfield network during the classification. MFA replaces the stochastic nature of simulated annealing method with a set of deterministic update rules that act on the average value of the variable. By minimizing averages, it is possible to converge to an equilibrium state considerably faster than standard simulated annealing method. Moreover, the proposed agglomerative clustering algorithm which determines the underlying clusters of the image provides initial input values of Hopfield neural network. Results : The proposed approach which uses agglomerative clustering and deterministic relaxation approach resolves the problem of local optimization and achieves fast convergency toward a global optimization when a neural network is used for MRI classification. Conclusion : In this paper, we introduce a new paradigm to classify MRI using clustering analysis and deterministic relaxation for neural network to improve the classification results.