• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.29 no.7B
• /
• pp.641-649
• /
• 2004

We propose a dynamic traffic management model which uses adaptive priority reassignment algorithm to deliver service differentiation in home networks, and implement adaptive priority reassignment algorithm using FPGA. The proposed architecture is designed to handle home network traffic without the need for signaling protocol. We categorize home network traffic into three kinds of traffic class: control data traffic class, the Internet data and non-real-time data traffic class, and multimedia data traffic class (include non-real-time and real-time multimedia data traffic). To support differential service about these kinds of traffic class, we designed and implemented a traffic management framework that dynamically change each traffic class priority depending on bandwidth utilization of each traffic class.

• The KIPS Transactions:PartC
• /
• v.11C no.2
• /
• pp.253-262
• /
• 2004

In this paper, a session allocation algorithm for a switch with multiple shared links is proposed. The algorithm guarantees the reserved bandwidth to each service class and keeps the delay of sessions belonging to a service class as close as possible even if the sessionsare allocated to different shared links. To support these qualities of services, a new scheduling model for multiple shared links is defined and a session allocation algorithm to decide a shared link to be allocated to a new session on the connection establishmentis developed based on the model. The proposed heuristic algorithm allocates a session to a link including the subclass with the shortest (expected) delay that subclasses of the service class the session belongs to will experience. Simulation results verify that a switch with multiple shared links hiring the proposed algorithm provides service classes with fairer bandwidth allocation and higher throughput, and guarantees reserved bandwidth better than the switch hiring other session algorithms. It also guarantees very similarservice delay to the sessions in the same service class.

• Journal of Intelligence and Information Systems
• /
• v.28 no.3
• /
• pp.23-43
• /
• 2022

In order to apply AI (Artificial Intelligence) in various industries, interest in algorithm selection is increasing. Algorithm selection is largely determined by the experience of a data scientist. However, in the case of an inexperienced data scientist, an algorithm is selected through meta-learning based on dataset characteristics. However, since the selection process is a black box, it was not possible to know on what basis the existing algorithm recommendation was derived. Accordingly, this study uses k-means cluster analysis to classify types according to data set characteristics, and to explore suitable classification algorithms and methods for resolving class imbalance. As a result of this study, four types were derived, and an appropriate class imbalance resolution method and classification algorithm were recommended according to the data set type.

• Journal of the Institute of Electronics Engineers of Korea TC
• /
• v.43 no.5 s.347
• /
• pp.94-103
• /
• 2006

Ethernet passive optical network (EPON) has drawn many attention as a promising access network technology for FTTH because it can provide a high bandwidth with a low cost. Since the uplink bandwidth in the EPON system is shared by many users, it is necessary for an EPON system to have an efficient bandwidth allocation mechanism. To support QoS in EPON, the previous bandwidth allocation schemes employ strict priority queueing (SPQ). Since SPQ gives unlimited priority to higher service class, the QoS of lower service classes gets worse. In this paper, we propose Class Gated DBA (Dynamic Bandwidth Allocation) algorithm in which the bandwidth is requested / granted in a service class basis. To avoid the monopoly in bandwidth usage by higher classes the maximum bandwidth that is allocate to each service class is limited (fairness between services classes). In addition, to avoid the monopoly in bandwidth usage by some particular users, each ONU runs fairness bandwidth allocation algorithm within each service classes. Through computer simulations, it is verified that the proposed algorithm achieves a good level of QoS, and at the same time maintains a good level of fairness between both service classes and users.

• Communications for Statistical Applications and Methods
• /
• v.27 no.1
• /
• pp.129-140
• /
• 2020

In this paper, we introduce an efficient algorithm for the non-convex penalized multinomial logistic regression that can be uniformly applied to a class of non-convex penalties. The class includes most non-convex penalties such as the smoothly clipped absolute deviation, minimax concave and bridge penalties. The algorithm is developed based on the concave-convex procedure and modified local quadratic approximation algorithm. However, usual quadratic approximation may slow down computational speed since the dimension of the Hessian matrix depends on the number of categories of the output variable. For this issue, we use a uniform bound of the Hessian matrix in the quadratic approximation. The algorithm is available from the R package ncpen developed by the authors. Numerical studies via simulations and real data sets are provided for illustration.

• Journal of information and communication convergence engineering
• /
• v.11 no.4
• /
• pp.258-267
• /
• 2013

New incremental learning algorithm using extended data expression, based on probabilistic compounding, is presented in this paper. Incremental learning algorithm generates an ensemble of weak classifiers and compounds these classifiers to a strong classifier, using a weighted majority voting, to improve classification performance. We introduce new probabilistic weighted majority voting founded on extended data expression. In this case class distribution of the output is used to compound classifiers. UChoo, a decision tree classifier for extended data expression, is used as a base classifier, as it allows obtaining extended output expression that defines class distribution of the output. Extended data expression and UChoo classifier are powerful techniques in classification and rule refinement problem. In this paper extended data expression is applied to obtain probabilistic results with probabilistic majority voting. To show performance advantages, new algorithm is compared with Learn++, an incremental ensemble-based algorithm.

• Journal of applied mathematics & informatics
• /
• v.32 no.1_2
• /
• pp.211-225
• /
• 2014

A new class of smoothing functions is introduced in this paper, which includes some important smoothing complementarity functions as its special cases. Based on this new smoothing function, we proposed a smoothing Newton method. Our algorithm needs only to solve one linear system of equations. Without requiring the nonemptyness and boundedness of the solution set, the proposed algorithm is proved to be globally convergent. Numerical results indicate that the smoothing Newton method based on the new proposed class of smoothing functions with ${\theta}{\in}(0,1)$ seems to have better numerical performance than those based on some other important smoothing functions, which also demonstrate that our algorithm is promising.

• The KIPS Transactions:PartA
• /
• v.15A no.3
• /
• pp.155-160
• /
• 2008

RFID technology has been gradually expanding its application. One of the important performance issues in RFID systems is to resolve the collision among multi-tags identification on restricted area. We consider a new anti-collision scheme based on Class Identification algorithm using Depth-First scheme. We evaluate how much performance can be improved by Class identification algorithm in the cases of Query-tree more then 17% identification rate and 150% performance.

• Journal of Communications and Networks
• /
• v.9 no.3
• /
• pp.312-318
• /
• 2007

In this paper, we propose the optimal offset-time decision (OOD) algorithm which can effectively reduce the pre-transmission delay compared to the perfect isolation model, and can also be extended to general n priority classes while the target loss probability of each class is guaranteed for the variable offered load. In order to drive the OOD algorithm, we first analyze the loss probability of each priority class through class aggregation and iteration method; the analytic results obtained through the proposed algorithm are then validated with results garnered from extensive simulation tests.

• Kyungpook Mathematical Journal
• /
• v.62 no.4
• /
• pp.729-736
• /
• 2022

To help solving intractable nonlinear evolution equations (NLEEs) of waves in the field of fluid dynamics we develop an algorithm to find new high order solutions of the class of Abel, Bernoulli, Chini and Riccati equations of the form y' = ayⁿ + by + c, n > 1, with constant coefficients a, b, c. The role of this class of equations in NLEEs is explained in the introduction below. The basic algorithm to compute the coefficients of the power series solutions of the class, emerged long ago and is further developed in this paper. Practical application for hitherto unknown solutions is exemplified.