• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.10 no.5
• /
• pp.293-298
• /
• 2010

The key of USN(Ubiquitous Sensor Network) technology is low power wireless communication technology and proper resource allocation technology for efficient routing. The distinguished resource allocation method is needed for efficient routing in sensor network. To solve this problems, we propose an algorithm that can be adopted in USN with making up for weak points of PQ and WRR in this paper. The proposed algorithm produces the control discipline by the fuzzy theory to dynamically assign the weight of WRR scheduler with checking the Queue status of each class in sensor network. From simulation results, the proposed algorithm improves the packet loss rate of the EF class traffic to 6.5% by comparison with WRR scheduling method and that of the AF4 class traffic to 45% by comparison with PQ scheduling method.

• Journal of the Institute of Electronics Engineers of Korea TC
• /
• v.40 no.6
• /
• pp.225-236
• /
• 2003

In this paper, we propose a wireless scheduling algorithm to provide the Internet delay proportional differentiated services in wireless networks. For considering network environments that have burst and location-dependent channel errors, our proposed WDPS(Wireless Delay Proportional Service) scheduling algorithm adaptively serves packets in class queues based on the delivered delay performance for each class. The remarkable characteristics of WDPS scheduler are supporting a fair relative delay service, providing graceful throughput and delay compensation, and avoiding class queue blocking problem. Through simulations, we show that the algorithm achieves the desirable properties to provide delay proportional services in wireless networks.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.9 no.8
• /
• pp.2894-2909
• /
• 2015

As the number of mobile devices such as smart phones and tablets explodes, the need for new services or applications is also rapidly increasing. Smart class application is one of the emerging applications, in which most of contents are distributed to all members of a class simultaneously. It is highly required to select relay nodes to cover shadow area of radio as well as extend coverage, but existing algorithms in a smart class environment suffer from high control packet overhead and delay for exchanging topology information among all pairs of nodes to select relay nodes. In addition, the relay selection procedure should be repeated in order to adapt to the dynamic topology changes caused by link status changes or device's movement. This paper proposes the learning based relay selection algorithm to overcome aforementioned problems. The key idea is that every node keeps track of its relay quality in a fully distributed manner, where RQI (Relay Quality Indicator) is newly defined to measure both the ability of receiving packets from content source and the ability of successfully relaying them to successors. The RQI of each node is updated whenever it receives or relays broadcast packet, and the node having the higher RQI is selected as a relay node in a distributed and run-time manner. Thus, the proposed algorithm not only removes the overhead for obtaining prior knowledge to select relay nodes, but also provides the adaptability to the dynamic topology changes. The network simulation and experimental results prove that the proposed algorithm provides efficient and reliable content distribution to all members in a smart class as well adaptability against network dynamics.

• Journal of applied mathematics & informatics
• /
• v.29 no.5_6
• /
• pp.1533-1539
• /
• 2011

In this paper, a class of nonlinear bilevel programs, i.e. the lower level problem is linear programs, is considered. Aiming at this special structure, we append the duality gap of the lower level problem to the upper level objective with a penalty and obtain a penalized problem. Using the penalty method, we give an existence theorem of solution and propose an algorithm. Then, a numerical example is given to illustrate the algorithm.

• Journal of applied mathematics & informatics
• /
• v.30 no.1_2
• /
• pp.127-133
• /
• 2012

We establish the polynomial convergence of a new class of path-following methods for SDLCP whose search directions belong to the class of directions introduced by Monteiro [3]. We show that the polynomial iteration-complexity bounds of the well known algorithms for linear programming, namely the short-step path-following algorithm of Kojima et al. and Monteiro and Alder, carry over to the context of SDLCP.

• 한국정보디스플레이학회:학술대회논문집
• /
• 2009.10a
• /
• pp.207-210
• /
• 2009

According to the recent demand for materials for use in various displays and solid state lightings, new phosphors with improved performance have been pursued consistently. Multi objective genetic algorithm assisted combinatorial material search (MOGACMS) strategies have been applied to various multi-compositional inorganic systems to search for new phosphors and to optimize the properties of phosphors.

• Proceedings of the KIEE Conference
• /
• 1996.07b
• /
• pp.1278-1280
• /
• 1996

In this paper, an Iterative learning control algorithm is presented for a class of non linear system which have external inputs or disturbances. The acceleration of error signal is used to update the next control signal. It is shown that the feedback gain can be deter.ined so that the overall errors are convergent.

• Information Systems Review
• /
• v.24 no.1
• /
• pp.39-57
• /
• 2022

This paper proposes the GMOPTBoost algorithm to improve the performance of the AdaBoost algorithm for bankruptcy prediction in which class imbalance problem is inherent. AdaBoost algorithm has the advantage of providing a robust learning opportunity for misclassified samples. However, there is a limitation in addressing class imbalance problem because the concept of arithmetic mean accuracy is embedded in AdaBoost algorithm. GMOPTBoost can optimize the geometric mean accuracy and effectively solve the category imbalance problem by applying Gaussian gradient descent. The samples are constructed according to the following two phases. First, five class imbalance datasets are constructed to verify the effect of the class imbalance problem on the performance of the prediction model and the performance improvement effect of GMOPTBoost. Second, class balanced data are constituted through data sampling techniques to verify the performance improvement effect of GMOPTBoost. The main results of 30 times of cross-validation analyzes are as follows. First, the class imbalance problem degrades the performance of ensembles. Second, GMOPTBoost contributes to performance improvements of AdaBoost ensembles trained on imbalanced datasets. Third, Data sampling techniques have a positive impact on performance improvement. Finally, GMOPTBoost contributes to significant performance improvement of AdaBoost ensembles trained on balanced datasets.

• Journal of the Korea Society of Computer and Information
• /
• v.10 no.2 s.34
• /
• pp.113-121
• /
• 2005

In this paper, we design the hybrid learning algorithm of LVQ which is to perform EMG pattern recognition. The proposed hybrid LVQ learning algorithm is the modified Counter Propagation Networks(C.p Net. ) which is use SOM to learn initial reference vectors and out-star learning algorithm to determine the class of the output neurons of LVa. The weights of the proposed C.p. Net. which is between input layer and subclass layer can be learned to determine initial reference vectors by using SOM algorithm and to learn reference vectors by using LVd algorithm, and pattern vectors is classified into subclasses by neurons which is being in the subclass layer, and the weights which is between subclass layer and class layer of C.p. Net. is learned to classify the classified subclass. which is enclosed a class . To classify the pattern vectors of EMG. the proposed algorithm is simulated with ones of the conventional LVQ, and it was a confirmation that the proposed learning method is more successful classification than the conventional LVQ.

• Nuclear Engineering and Technology
• /
• v.54 no.12
• /
• pp.4684-4692
• /
• 2022

Radionuclide identification is an important part of the nuclear material identification system. The development of artificial intelligence and machine learning has made nuclide identification rapid and automatic. However, many methods directly use existing deep learning models to analyze the gamma-ray spectrum, which lacks interpretability for researchers. This study proposes an explainable radionuclide identification algorithm based on the convolutional neural network and class activation mapping. This method shows the area of interest of the neural network on the gamma-ray spectrum by generating a class activation map. We analyzed the class activation map of the gamma-ray spectrum of different types, different gross counts, and different signal-to-noise ratios. The results show that the convolutional neural network attempted to learn the relationship between the input gamma-ray spectrum and the nuclide type, and could identify the nuclide based on the photoelectric peak and Compton edge. Furthermore, the results explain why the neural network could identify gamma-ray spectra with low counts and low signal-to-noise ratios. Thus, the findings improve researchers' confidence in the ability of neural networks to identify nuclides and promote the application of artificial intelligence methods in the field of nuclide identification.