• Journal of the Korea Society of Computer and Information
• /
• v.15 no.7
• /
• pp.11-17
• /
• 2010

This paper proposes an improved SMO solving a quadratic optmization problem for class imbalanced learning. The SMO algorithm is aproporiate for solving the optimization problem of a support vector machine that assigns the different regularization values to the two classes, and the prosoposed SMO learning algorithm iterates the learning steps to find the current optimal solutions of only two Lagrange variables selected per class. The proposed algorithm is tested with the UCI benchmarking problems and compared to the experimental results of the SMO algorithm with the g-mean measure that considers class imbalanced distribution for gerneralization performance. In comparison to the SMO algorithm, the proposed algorithm is effective to improve the prediction rate of the minority class data and could shorthen the training time.

• The Journal of the Korea Contents Association
• /
• v.3 no.2
• /
• pp.39-45
• /
• 2003

Dynamic dan loading and class linking of Java is a poweful mechanism. Many other system also support some form of dynamic loading and linking, but lazy loading, type-safe linkage, user-definable class loading policy, and multiple namespaces are important features of Java The ue U dan loading is assured of type safety. The security of Java greatly depends on type safety. In JVM, type safety mechanism is very difficult and access of accuracy is not dear, so type safety problems were raised. In paper, n analysis simple Java code and present a diagram graph and an operational semantics for dynamic class loading and type safety.

• The KIPS Transactions:PartD
• /
• v.8D no.1
• /
• pp.71-80
• /
• 2001

Object-Oriented (OO) methodology to use the concept like encapsulation, inheritance, polymorphism, and message passing demands metrics that are different from structured methodology. There are many studies for OO software metrics such as program complexity or design metrics. But the metrics for the analysis class need to decrease the complexity in the analysis phase so that greatly reduce the effort and the cost of system development. In this paper, we propose new metrics to measure the complexity of analysis classes which draw out in the analysis phase based on Unified Process. By the collaboration complexity, is denoted by CC, we mean the maximum number of the collaborations can be achieved with each of the collaborator and detennine the potential complexity. And the interface complexity, is denoted by IC, shows the difficulty related to understand the interface of collaborators each other. We prove mathematically that the suggested metrics satisfy OO characteristics such as class size and inheritance. And we verify it theoretically for Weyuker' s nine properties. Moreover, we show the computation results for analysis classes of the system which automatically respond to questions of the it's user using the text mining technique. As we compared CC and IC to CBO and WMC, the complexity can be represented by CC and IC more than CBO and WMC. We expect to develop the cost-effective OO software by reviewing the complexity of analysis classes in the first stage of SDLC (Software Development Life Cycle).

• Journal of KIISE:Computing Practices and Letters
• /
• v.14 no.8
• /
• pp.783-787
• /
• 2008

IEEE 802.16/WiBro, one of the representative Broadband Wireless Access (BWA) systems, defines QoS negotiation between BS and SS and five service classes to guarantee QoS for flows. The five service classes are UGS, ertPS, rtPS, nrtPS, and BE, but the standard does not define how to provide services to flows according to the service classes. Existing research works have been studied about the scheduler architecture to effectively utilize the IEEE 802.16/WiBro systems. These works use the original packet scheduling algorithm for service classes. However, it is necessary to consider bandwidth allocation mechanisms in scheduling algorithms since bandwidth allocation mechanisms are different according to service classes. Especially, bandwidth allocation mechanisms should be considered for the scheduling of rtPS class since rtPS class uses the polling mechanism that takes time to allocate bandwidth and has the minimum latency constraint. Therefore, we propose two stage dropping scheme for rtPS class that is sensitive to latency, and thus it reduces the wasted resources and provides efficient service to rtPS class.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.41 no.9
• /
• pp.1026-1037
• /
• 2016

3rd Generation Partnership Project (3GPP) is the progressing standardization of the narrowband IoT (NB-IoT) system to support massive devices for the Internet of Things (IoT) services. The NB-IoT system uses a coverage class technique to increase the performance of the NB-IoT system while serving massive devices in very wide coverage area. A moving device can change the coverage class according to the distance or the channel state between the base station and the moving device. However, in the conventional NB-IoT standard, the performance of the NB-IoT system degrades because the coverage class is changed based on the fixed criterion. This paper proposes the coverage class adaptation schemes to increase the performance of the NB-IoT system by dynamically change the coverage class according to the location or the channel state of the device. Simulation results show that the proposed coverage class adaptation scheme decreases both the signaling overhead and the PDCCH decoding error rate in comparison with the conventional coverage class adaptation scheme in the 3GPP standard.

• The Journal of Korea Institute of Information, Electronics, and Communication Technology
• /
• v.12 no.4
• /
• pp.406-412
• /
• 2019

In this paper, we propose two algorithms for classifying services with similar delay characteristics into three classes and allocating radio resources according to priority in LTE-A system where H2H and M2M services coexist. The first is to allocate resources from the higher priority class to the lower priority class, and each class gives priority to H2H over M2M, and the other is to give priority to H2H regardless of delay characteristics except for the class with the highest priority. The RA success probability was analyzed according to the access rate(${\alpha}$) of M2M devices in each class. In comparison with the conventional systems, it was improved from 0.5 to 0.52 for ${\alpha}_{2M}=0.05$ in two classes. In the three classes, the success probability was slightly increased from 0.5 to 0.57 for ${\alpha}_{2M}={\alpha}_{3M}=1$ and from 0.5 to 0.58 for ${\alpha}_{2M}=0.5$ and ${\alpha}_{3M}=0.1$. Although 6 services are considered in the proposed scheme, the RA success probability is almost similar to the previous scheme because the average arrival rate of H2H of each class is set to the same.

• Journal of KIISE:Software and Applications
• /
• v.26 no.8
• /
• pp.1028-1038
• /
• 1999

본 논문에선 LR 문법의 부분 클래스를 동치인 LL 문법으로 변환하는 방법을 제시한다. 이 변환이 적용 가능한 문법을 확장된k`-transformable 문법이라 정의한다. 변환된 문법은 left-to-right 커버의 성질을 만족한다. 기존 연구에서 제시한 변환 방법은 LR 문법의 부분 클래스인 {{{{k`-transformable 클래스와 PLR 클래스를 LL 문법으로 바꾼다. 이 논문에서 제시하는 새로운 변환 방법의 적용 가능한 문법의 범위는 k`-transformable 클래스와 PLR 클래스를 포함한다. 기존의 커버링 성질을 만족하는 LL로의 문법 변환은 보편적인 LR 파서의 행동을 시뮬레이션하여 얻어진다. 이 과정에서 쉬프트, 리듀스 행동 이외에 무한의 가능성을 가진 스택 스트링의 유한 표현을 위해 리덕션 심볼에 대한 예상 행동이 추가된다. 본 논문에서는 파싱 문맥을 나타내는 LR 아이템들을 기존의 스택 스트링 표현 형태에 추가하여 스택 스트링 표현법을 정제하고, 리덕션 심볼에 대한 예상 방법을 확장하는 정형식을 제시한다. 이에 근거하여 LL 커버링 문법이 존재하는 클래스를 확장된 {{{{k`-transformable 문법으로 확장시킨다.Abstract A new transformation of a subclass of LR(k`) grammars into equivalent LL(k`) grammars is studied. The subclass of LR(k`) grammars is called extended k`-transformable. The transformed LL(k`) grammars left-to-right cover the original LR(k`) grammars. Previous transformations transform k`-transformable and PLR(k`) into LL(k`). The new transformation is more powerful in that it handles the extended k`-transformable subclass of LR(k`), which strictly includes k`-transformable and PLR(k`) classes. The previous covering transformations into LL grammars are obtained by simulating the actions of the conventional LR parser. Specially, a predict action of reduction goals is added to the action set in order to finitely represent stack string. In this paper, the stack string representation is refined by adding LR items to represent a parsing context, and the prediction of reduction goal is extended by generalizing the prediction formalism. Based on them, the previous grammar classes with LL({{{{k`) covering grammars are extended to extended k`-transformable grammars.

• Journal of the Computational Structural Engineering Institute of Korea
• /
• v.28 no.6
• /
• pp.579-588
• /
• 2015

The data model for structural experiment information formally organizes the information involved in the structural experiments before the data repository using the data model is implemented. The data model is particularly required for the data repositories for the large-scale structural experiment information and the general information for various types of experiments, such as the NEEShub Project Warehouse developed by NEES. This paper proposes criteria for evaluating the organization and the use of design model for structural experiment information. The term of AVE(attribute value existence) indicates the ratio of attributes who values exist in objects, and then used for defining the Attribute AVE for the use of an attribute, the Class AVE for a class, the Class Level AVE for a class including its lower-level classes, the Project AVE for a project including all classes at class levels, and the Data Model AVE for a data model including projects. These criteria are applied to the projects in the NEES data model, and it is successively possible to numerically describe the evaluation of the use of classes and attributes in the data model.

• The Journal of the Korea Contents Association
• /
• v.6 no.5
• /
• pp.85-95
• /
• 2006

In this paper, we propose a method for simplifying the modeling of 'walk' for various instances of quadruped and easily extending the model of 'walk' for the instance of new subordinate class which is not defined in the original class hierarchy by using the sophisticatedly modeled 'walk' of quadruped. To achieve this method, we apply the analyzed pattern of walk and classification according to the structural similarity of quadruped studied in the previous researches to the class hierarchy and propose a method for inheriting the actions of super class. This paper model the 'walk' of quadruped by concertizing the characteristics of quadruped and defining the necessary factors and appropriate domains in terms of parameters of 'walk' and apply the model to the horse and cow, typical instances of quadruped.

• Journal of KIISE:Computer Systems and Theory
• /
• v.35 no.8
• /
• pp.372-377
• /
• 2008

Multiclass cancer classification has been actively investigated based on gene expression profiles, where it determines the type of cancer by analyzing the large amount of gene expression data collected by the DNA microarray technology. Since gene expression data include many genes not related to a target cancer, it is required to select informative genes in order to obtain highly accurate classification. Conventional rank-based gene selection methods often use ideal marker genes basically devised for binary classification, so it is difficult to directly apply them to multiclass classification. In this paper, we propose a novel method for multiclass gene selection, which does not use ideal marker genes but directly analyzes the distribution of gene expression. It measures the class-discriminability by discretizing gene expression levels into several regions and analyzing the frequency of training samples for each region, and then classifies samples by using the naive Bayes classifier. We have demonstrated the usefulness of the proposed method for various representative benchmark datasets of multiclass cancer classification.