• Journal of Soil and Groundwater Environment
• /
• v.20 no.3
• /
• pp.51-64
• /
• 2015

In the present study, we develop two history matching techniques based on Markov chain Monte Carlo method where radial basis function and Gaussian distribution generated by unconditional geostatistical simulation are employed as the random walk transition kernels. The Bayesian inverse methods for aquifer characterization as the developed models can be effectively applied to the condition even when the targeted information such as hydraulic conductivity is absent and there are transient hydraulic head records due to imposed stress at observation wells. The model which uses unconditional simulation as random walk transition kernel has advantage in that spatial statistics can be directly associated with the predictions. The model using radial basis function network shares the same advantages as the model with unconditional simulation, yet the radial basis function network based the model does not require external geostatistical techniques. Also, by employing radial basis function as transition kernel, multi-scale nested structures can be rigorously addressed. In the validations of the developed models, the overall predictabilities of both models are sound by showing high correlation coefficient between the reference and the predicted. In terms of the model performance, the model with radial basis function network has higher error reduction rate and computational efficiency than with unconditional geostatistical simulation.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.29 no.12C
• /
• pp.1660-1668
• /
• 2004

This paper proposes a Voice Activity Detection (VAD) algorithm based on Radial Basis Function (RBF) network using dual threshold. The k-means clustering and Least Mean Square (LMS) algorithm are used to upade the RBF network to the underlying speech condition. The inputs for RBF are the three parameters in a Code Exited Linear Prediction (CELP) coder, which works stably under various background noise levels. Dual hangover threshold applies in BRF-VAD for reducing error, because threshold value has trade off effect in VAD decision. The experimental result show that the proposed VAD algorithm achieves better performance than G.729 Annex B at any noise level.

• 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%.

• Journal of KIISE:Software and Applications
• /
• v.27 no.9
• /
• pp.913-919
• /
• 2000

본 논문에서는 인간 시지각을 고려한 새로운 영상 압축 방법을 제시한다. 영상의 화소의 값들이 x-y 평명상에서 정의된 3차원 곡면 위에 있는 점들로 가정하여, 영상을 곡면의 복잡도에 따라 나누고, 나누어진 각각의 곡면(영역)은 Radial-Basis Function (RBF)를 사용하여 근사화하는 방법으로 영상을 압축한다. 본 방법은 JPEG 압축 방법과 비슷한 압축율과 영상의 질을 얻을 수 있다.

• 한국전산유체공학회:학술대회논문집
• /
• 2009.04a
• /
• pp.121-124
• /
• 2009

A moving mesh system is one of the critical parts in a computational fluid dynamics analysis. In this study, the RBF(Radial Basis Function) which shows better performance than hybrid meshes was developed to obtain the deformed grid. The RBF method can handle large mesh deformations caused by translations, rotations and deformations, both for 2D and 3D meshes. Another advantage of the method is that it can handle both structured and unstructured grids with ease. The method uses a volume spline technique to compute the deformation of block vertices and block edges, and deformed shape.

• Proceedings of the KIEE Conference
• /
• 2008.10b
• /
• pp.237-238
• /
• 2008

퍼지 RBFNN의 설계에 있어 가장 중요한 과정인 Radial Basis Function의 결정은 퍼지 RBFNN의 모델링 성능을 좌우한다. 기존에는 FCM을 이용하여 Radial Basis Function의 초기 위치를 결정하고 오류 역전파 알고리즘과 같은 최적화 알고리즘을 이용하여 최적의 Radial Basis Function을 결정하였다. 근래에는 Conditional FCM을 이용하여 출력공간에 정의된 정보입자의 정보를 이용하여 입력공간상에서 Radial Basis Function의 위치를 결정하여 퍼지 RBFNN의 성능을 개선시키고자 하는 연구 수행되어졌다. 그러나 출력공간상에서 얻은 정보입자를 입력공간상으로 정보 손실없이 전달할 수 없어서 기대한 만큼의 성능 개선을 이룰 수 없었다. 이를 개선하기 위해 출력 공간예서 정의된 정보 입자를 정보 손실없이 입력 공간에 투영하기 위하여 퍼지 kNN기법을 도입하여 새로운 퍼지 RBFNN 설계 방법을 제안한다.

• 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.

• Journal of Electrical Engineering and Technology
• /
• v.11 no.6
• /
• pp.1872-1879
• /
• 2016

The techniques to recycle and reuse plastics attract public attention. These public attraction and needs result in improving the recycling technique. However, the identification technique for black plastic wastes still have big problem that the spectrum extracted from near infrared radiation spectroscopy is not clear and is contaminated by noise. To overcome this problem, we apply Raman spectroscopy to extract a clear spectrum of plastic material. In addition, to improve the classification ability of fuzzy Radial Basis Function Neural Networks, we apply supervised learning based clustering method instead of unsupervised clustering method. The conditional fuzzy C-Means clustering method, which is a kind of supervised learning based clustering algorithms, is used to determine the location of radial basis functions. The conditional fuzzy C-Means clustering analyzes the data distribution over input space under the supervision of auxiliary information. The auxiliary information is defined by using k Nearest Neighbor approach.

• 제어로봇시스템학회:학술대회논문집
• /
• 1996.10b
• /
• pp.1165-1168
• /
• 1996

The neural network approach has been shown to be a general scheme for nonlinear dynamical system identification. Unfortunately the error surface of a Multilayer Neural Network(MNN) that widely used is often highly complex. This is a disadvantage and potential traps may exist in the identification procedure. The objective of this paper is to identify a nonlinear dynamical systems based on Radial Basis Function Networks(RBFN). The learning with RBFN is fast and precise. This paper discusses RBFN as identification procedure is based on a nonlinear dynamical systems. and A design method of model follow control system based on RBFN controller is developed. As a result of applying this method to inverted pendulum, the simulation has shown that RBFN can be used as identification and control of nonlinear dynamical systems effectively.

• Proceedings of the Acoustical Society of Korea Conference
• /
• 1997.06a
• /
• pp.13-16
• /
• 1997

본 논문에서는 Bayesian 결정 이론을 이용한 기존의 Radial Basis Function Networks 이되는 출력층에서 선형 조합되는 것과는 다른 형태의 방법을 제안하고자 한다. 제안하고자 하는 방법은 은닉층의 출력값과 가중치와의 곱해진 값이 출력층의 입력으로 들어오는데 이들 입력신호를 경쟁을 통하여 가장 큰 값만을 출력신호 인정하는 방법이다. 이런 경우에 파라미터 갱신을 할 때도 모든 가중치를 다 갱신하는 것이 아니라 출력되는 은닉층에 연결된 가중치만을 갱신하게된다. 이렇게 할 경우 계산량 감소뿐만 아니라 학습시간을 단축할 수 있다는 장점이 있다. 그리고 제안한 방법을 이용할 경우 비선형 분류문제에서도 우수한 성능결과를 확인 할 수 있었으며 기존의 RBFN rhk Wiener Filter와 성능을 비교하였다.