• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.6 no.5
• /
• pp.1333-1353
• /
• 2012

The robustness of a network is usually measured by error tolerance and attack vulnerability. Significant research effort has been devoted to determining the network design with optimal robustness. However, little attention has been paid to the problem of how to improve the robustness of existing networks. In this paper, we investigate how to optimize attack tolerance and communication efficiency of an existing network under the limited link addition. A survival fitness metric is defined to measure both the attack tolerance and the communication efficiency of the network. We show that network topology reconfiguration optimization with limited link addition (NTRLA) problem is NP-hard. Two approximate solution methods are developed. First, we present a degree-fitness parameter to guide degree-based link addition method. Second, a preferential configuration node-protecting cycle (PCNC) method is developed to do trade-off between network robustness and efficiency. The performance of PCNC method is demonstrated by numerical experiments.

• International Journal of Control, Automation, and Systems
• /
• v.6 no.1
• /
• pp.24-34
• /
• 2008

A novel neural network model, termed the standard neural network model (SNNM), similar to the nominal model in linear robust control theory, is suggested to facilitate the synthesis of controllers for delayed (or non-delayed) nonlinear systems composed of neural networks. The model is composed of a linear dynamic system and a bounded static delayed (or non-delayed) nonlinear operator. Based on the global asymptotic stability analysis of SNNMs, Static state-feedback controller and dynamic output feedback controller are designed for the SNNMs to stabilize the closed-loop systems, respectively. The control design equations are shown to be a set of linear matrix inequalities (LMIs) which can be easily solved by various convex optimization algorithms to determine the control signals. Most neural-network-based nonlinear systems with time delays or without time delays can be transformed into the SNNMs for controller synthesis in a unified way. Two application examples are given where the SNNMs are employed to synthesize the feedback stabilizing controllers for an SISO nonlinear system modeled by the neural network, and for a chaotic neural network, respectively. Through these examples, it is demonstrated that the SNNM not only makes controller synthesis of neural-network-based systems much easier, but also provides a new approach to the synthesis of the controllers for the other type of nonlinear systems.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.20 no.1
• /
• pp.155-161
• /
• 2020

In general, optimization problems are difficult to solve simply. The reason is that the given problem is solved as soon as it is simple, but the more complex it is, the very large number of cases. This study is about the optimization of AI neural network. What we are dealing with here is the relief method for constructing AI network. The main topics deal with non-deterministic issues such as the stability and unstability of the overall network state, cost down and energy down. For this one, we discuss associative memory models, that is, a method in which local minimum memory information does not select fake information. The simulated annealing, this is a method of estimating the direction with the lowest possible value and combining it with the previous one to modify it to a lower value. And nonlinear planning problems, it is a method of checking and correcting the input / output by applying the appropriate gradient descent method to minimize the very large number of objective functions. This research suggests a useful approach to relief method as a theoretical approach to solving optimization problems. Therefore, this research will be a good proposal to apply efficiently when constructing a new AI neural network.

• Journal of the Institute of Electronics and Information Engineers
• /
• v.53 no.4
• /
• pp.27-34
• /
• 2016

In the train system, railway vehicles are connected in a line. Therefore, this feature should be considered in composing network topology in a train system. Besides, inter-car communication should be distinguished from in-car communication. As for the inter-car communication, the hybrid topology was proposed to use rather than the conventional ring, star, daisy-chain, and bus topologies. In the hybrid topology, a number of cars are bound to be a group. Then star topology is used for the communication in a group and daisy-chain topology is used for the communication between groups. Hybrid topology takes the virtue of both star and daisy-chain topologies. Hence it maintains communication speed with reducing the number of connecting cables between cars. Therefore, it is important to choose the number of cars in a group to obtain higher performance. In this paper, we focus on the optimization of hybrid topology for railway cars. We first assume that the size of data and the frequency of data production for each car is identical. We also assume that the importance for the maximum number of cables to connect cars is variable as well as the importance of the communication speed. Separated weights are granted to both importance and we derive the optimum number of cars in a group for various number of cars and weights.

• Journal of The Korean Association of Information Education
• /
• v.25 no.5
• /
• pp.791-801
• /
• 2021

In this paper, an educational case that can be applied as artificial intelligence literacy education for preservice teachers and incumbent teachers was studied. To this end, a case of educating the operating principle of an artificial neural network that recognizes images is proposed. This training case focuses on the basic principles of artificial neural network operation and implementation, and applies the method of finding parameter optimization solutions required for artificial neural network implementation in a spreadsheet. In this paper, we focused on the artificial neural network of supervised learning method. First, as an artificial neural network principle education case, an artificial neural network education case for recognizing two types of images was proposed. Second, as an artificial neural network extension education case, an artificial neural network education case for recognizing three types of images was proposed. Finally, the results of analyzing artificial neural network training cases and training satisfaction analysis results are presented. Through the proposed training case, it is possible to learn about the operation principle of artificial neural networks, the method of writing training data, the number of parameter calculations executed according to the amount of training data, and parameter optimization. The results of the education satisfaction survey for preservice teachers and incumbent teachers showed a positive response result of over 70% for each survey item, indicating high class application suitability.

• 제어로봇시스템학회:학술대회논문집
• /
• 1996.10a
• /
• pp.235-238
• /
• 1996

In this paper, a new optimization method which is a kind of random searching is presented. The proposed method is called RasVal which is an abbreviation of Random Search Method with Variable Seaxch Length and it can search for a global minimum based on the probability density functions of searching, which can be modified using informations on success or failure of the past searching in order to execute intensified and diversified searching. By applying the proposed method to a nonlinear crane control system which can be controlled by the Universal Learning Network with radial basis function(R.B.P.), it has been proved that RasVal is superior in performance to the commonly used back propagation learning algorithm.

• 제어로봇시스템학회:학술대회논문집
• /
• 2003.10a
• /
• pp.1982-1985
• /
• 2003

Future cellular networks will mainly be driven by, high quality channels, high band with utilization, low power consumption and efficient network management. For a given channel allocation, the capacity and quality of communication of cellular radio systems using CDMA(Code Division Multiple Access) can be increased by using a transmitter power control scheme to combat the near-far problem. Centralized power control schemes or distributed ones to maximize the minimum signal-to-interference in each user of CDMA wireless network have been investigated. This paper has proposed a distributed power control algorithm, which employs an adaptive search scheme, in order to solve quickly the linear systems of equations for power update in CDMA cellular radio systems. The simulation results show that the proposed scheme has faster convergence rate than the typical bang-bang type of distributed power control algorithm, which has been much used as a reference algorithm in IS-95A and W-CDMA communication network.

• IEIE Transactions on Smart Processing and Computing
• /
• v.6 no.3
• /
• pp.193-199
• /
• 2017

Instead of using only spatial features from a single frame for person re-identification, a combination of spatial and temporal factors boosts the performance of the system. A recurrent neural network (RNN) shows its effectiveness in generating highly discriminative sequence-level human representations. In this work, we implement RNN, three Long Short Term Memory (LSTM) network variants, and Gated Recurrent Unit (GRU) on Caffe deep learning framework, and we then conduct experiments to compare performance in terms of size and accuracy for person re-identification. We propose using GRU for the optimized choice as the experimental results show that the GRU achieves the highest accuracy despite having fewer parameters than the others.

• International journal of advanced smart convergence
• /
• v.6 no.4
• /
• pp.73-79
• /
• 2017

The Convolutional Neural Network (CNN) has shown an excellent performance in computer vision task. Applications of CNN include image classification, object detection in images, autonomous driving, etc. This paper will evaluate the performance of CNN model with ReLU and SELU as activation function. The evaluation will be performed on four different choices of hyperparameter which are initialization method, network configuration, optimization technique, and regularization. We did experiment on each choice of hyperparameter and show how it influences the network convergence and test accuracy. In this experiment, we also discover performance improvement when using SELU as activation function over ReLU.

• 제어로봇시스템학회:학술대회논문집
• /
• 1993.10b
• /
• pp.76-80
• /
• 1993

In this paper we discuss optimization of neural network parameters, such as inclination of the sigmoid function, the numbers of the input layer's units and the hidden layer's units, considering application to recognition of hand written Hangul. Hangul characters are composed of vowels and consonants, and basically classified to six patterns by their positions. Using these characteristics of Hangul, the pattern of a given character is determined by its peripheral distribution and the other features. After then, the vowels and the consonants are recognized by the optimized neural network. The constructed recognition system including a neural network is applied to non-learning Hangul written by some Korean people, which are the names randomly taken from Korean spiritual and cultural research institute.