• Proceedings of the Korean Society of Computer Information Conference
• /
• 2008.06a
• /
• pp.149-154
• /
• 2008

FCM 기반 RBF 네트워크는 서로 다른 학습 구조가 결합된 혼합형 모델로서, 입력층과 중간층의 학습 구조는 FCM 알고리즘을 적용하고, 중간층과 출력층 사이의 학습 구조는 Max_Min 신경망을 적용한다. 입력층과 중간층의 학습시 입력벡터와 중간층의 노드중에서 중심과 입력벡터간의 가장 가까운 노드를 승자 노드로 선택하여 출력층으로 전달한다. 그리고 중간층과 출력층 사이의 학습 구조는 Max_Min 신경망을 적용하여 중간층의 승자 뉴런이 출력층의 입력벡터로 적용한다. 하지만 많은 패턴이 입력벡터로 제시될 경우 학습 성능이 저하되는 단점이 있다. 따라서 본 논문에서는 중간층과 출력층의 학습 구조인 Max_Min 알고리즘의 학습 성능을 개선시키기 위해 퍼지 제어시스템을 이용하여 학습률을 동적으로 조정하는 퍼지 제어 기법을 이용한 FCM 기반 RBF 네트워크를 제안한다. 제안된 방법의 학습 성능을 평가하기 위하여 컨테이너 영상에서 추출한 숫자, 영문 식별자를 학습 데이터로 적용한 결과, 기존의 ART2 기반 RBF 네트워크보다 학습 시간이 적게 소요되고, 학습의 수렴성이 개선된 것을 확인하였다.

• Journal of the Korea Society of Computer and Information
• /
• v.19 no.9
• /
• pp.33-38
• /
• 2014

The FCM based hybrid RBF network is a heterogeneous learning network model that applies FCM algorithm between input and middle layer and applies Max_Min algorithm between middle layer and output. The Max-Min neural network uses winner nodes of the middle layer as input but shows inefficient learning in performance when the input vector consists of too many patterns. To overcome this problem, we propose a dynamic learning rate control based on fuzzy logic. The proposed method first classifies accurate/inaccurate class with respect to the difference between target value and output value with threshold and then fuzzy membership function and fuzzy decision logic is designed to control the learning rate dynamically. We apply this proposed RBF network to the character recognition problem and the efficacy of the proposed method is verified in the experiment.

• Journal of the Institute of Electronics Engineers of Korea SC
• /
• v.37 no.5
• /
• pp.46-55
• /
• 2000

This paper presents a direct adaptive control of robot system using chaotic neural networks and PD controller. The chaotic neural networks have robust nonlinear dynamic characteristics because of the sufficient nonlinearity in neuron itself, and the additional self-feedback and inter-connecting weights between neurons in same layer. Since the structure and the learning method are not appropriate for applying in control system, this neural networks have not been applied. In this paper, a modified chaotic neural networks is presented for dynamic control system. To evaluate the performance of the proposed neural networks, these networks are applied to the trajectory control of the three-axis PUMA robot. The structure of controller consists of PD controller and chaotic neural networks in parallel for conforming the stability in initial learning phase. Therefore, the chaotic neural network controller acts as a compensating controller of PD controller.

• Journal of the Institute of Convergence Signal Processing
• /
• v.2 no.1
• /
• pp.44-52
• /
• 2001

This paper describes an off-line method for recognizing totally unconstrained handwritten digits using modified chaotic neural networks(MCNN). The chaotic neural networks(CNN) is modified to be a useful network for solving complex pattern problems by enforcing dynamic characteristics and learning process. Since the MCNN has the characteristics of highly nonlinear dynamics in structure and neuron itself, it can be an appropriate network for the robust classification of complex handwritten digits. Digit identification starts with extraction of features from the raw digit images and then recognizes digits using the MCNN based classifier. The performance of the MCNN classifier is evaluated on the numeral database of Concordia University, Montreal, Canada. For the relative comparison of recognition performance, the MCNN classifier is compared with the recurrent neural networks(RNN) classifier. Experimental results show that the classification rate is 98.0%. It indicates that the MCNN classifier outperforms the RNN classifier as well as other classifiers that have been reported on the same database.

• Proceedings of the Korean Society of Precision Engineering Conference
• /
• 1996.11a
• /
• pp.759-763
• /
• 1996

This paper presents a new approach to the design of neural control system using digital signal processors in order to improve the precision and robustness. Robotic manipulators have become increasingly important in the field of flexible automation. High speed and high-precision trajectory tracking are indispensable capabilities for their versatile application. The need to meet demanding control requirement in increasingly complex dynamical control systems under significant uncertainties, leads toward design of intelligent manipulation robots. The TMS320C31 is used in implementing real time neural control to provide an enhanced motion control for robotic manipulators. In this control scheme, the networks introduced are neural nets with dynamic neurons, whose dynamics are distributed over all the network nodes. The nets are trained by the distributed dynamic back propagation algorithm. The proposed neural network control scheme is simple in structure, fast in computation, and suitable for implementation of real-time control. Performance of the neural controller is illustrated by simulation and experimental results for a SCARA robot.

• The Korean Journal of Applied Statistics
• /
• v.36 no.5
• /
• pp.361-380
• /
• 2023

Magnetoencephalography (MEG) is a technique to record oscillatory magnetic fields coming from ongoing neuronal activity. Functional brain activities performing cognitive or physiological tasks are performed on structural connections between neurons or brain regions. MEG data can be characterized as highly correlated, spatio-temporal, multidimensional, multilayered dynamic networks. Due to its complex structure, many studies on MEG network have not yet been conducted. In this study, we will explain the concept, necessity, and possible approaches of MEG network analysis. We reviewed the characteristics of MEG data. Network measures and potential network models in MEG and clinical studies are also reviewed.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2008.05a
• /
• pp.359-365
• /
• 2008

CP(Counterpropagation)알고리즘은 서로 다른 두 개의 신경망이 하나로 결합 된 혼합형 모델로서, 다른 신경망 모델에 비해 비교적 단순하고 빠른 학습 속도를 보인다. 그러나 CP 알고리즘은 다양한 패턴이 입력되면 충분한 경쟁층의 수가 설정되지 않아 학습이 불안정하고, 출력층에서 연결강도를 조정할 때 일반적인 학습률 조정방법으로 불안정한 학습 결과를 보인다. 이러한 문제점을 해결하기 위해 다수의 경쟁층을 설정하여 경쟁층에서 패턴 분류의 정확성을 높이고, 입력 벡터와 승자 뉴런의 대표 벡터간의 차이와 승자 빈도수를 반영하여 학습률을 동적으로 조정하여 경쟁층에서의 학습이 안정적으로 진행되도록 하고, 출력층에서 연결강도를 조정할 때 모멘텀(momentum)학습법을 적용한 개선된 CP 알고리즘이 제안되었다. 본 논문에서는 개선된 CP 알고리즘에서 경쟁층의 수를 효율적으로 설정하기 위해 퍼지 제어 기법을 이용하여 경쟁층의 수를 결정하는 방법을 제안한다. 제안된 방법은 CP 알고리즘에 입력되는 패턴의 정보를 이용하여 퍼지 소속 함수를 설계하고 입력에 대한 소속도를 계산한 후, 퍼지 제어 규칙을 적용하고, Mamdani의 Min_Max 추론 방법으로 추론한다. 퍼지 추론을 통해 최종적으로 얻어진 값을 무게 중심법으로 비퍼지화 하여 최종적으로 개선된 CP 알고리즘의 경쟁층의 수를 결정하는데 적용한다. 제안된 방법의 학습 및 인식 성능을 평가하기 위해, 숫자, 영어 등과 같이 다양한 패턴을 실험에 적용한 결과, 제안된 방법이 경쟁층의 수를 결정하는데 효과적임을 확인할 수 있었다.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.10 no.5
• /
• pp.215-221
• /
• 2010

Hopfield networks have been proposed as a new computational tool for finding the shortest path of networks. Zhang and Ali studied the method of finding shortest path by expended neurons of Hopfield networks. Ali Algorithm is well known as the tool with the neurons of branch numbers. Where a network grows bigger, it needs much more time to solve the problem by Ali algorithm. This paper modifies the method to find the synapse matrix and the input bias vector. And it includes the eSPN algorithm after proper iterations of the Hopfield network. The proposed method is a tow-stage method and it is more efficient to find the shortest path.The proposed method is verified by three sample networks. And it could be more applicable then Ali algorithm because it's fast and easy. When the cost of brach is changed, the proposed method works properly. Therefore dynamic cost-varing networks could be used by the proposed method.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.22 no.2
• /
• pp.135-140
• /
• 2012

The mirror neuron system in the cerebrum, which are handled by visual information-based imitative learning. When we observe the observer's range of mirror neuron system, we can assume intention of performance through progress of neural activation as specific range, in include of partially hidden range. It is goal of our paper that imitative learning is applied to 3D vision-based intelligent system. We have experiment as stereo camera-based restoration about acquired 3D image our previous research Using Optical flow, unscented Kalman filter. At this point, 3D input image is sequential continuous image as including of partially hidden range. We used Hidden Markov Model to perform the intention recognition about performance as result of restoration-based hidden range. The dynamic inference function about sequential input data have compatible properties such as hand gesture recognition include of hidden range. In this paper, for proposed intention recognition, we already had a simulation about object outline and feature extraction in the previous research, we generated temporal continuous feature vector about feature extraction and when we apply to Hidden Markov Model, make a result of simulation about hand gesture classification according to intention pattern. We got the result of hand gesture classification as value of posterior probability, and proved the accuracy outstandingness through the result.