• Proceedings of the Korea Inteligent Information System Society Conference
• /
• 1999.10a
• /
• pp.365-373
• /
• 1999

Recently, numerous studies have demonstrated that artificial intelligence such as neural networks can be an alternative methodology for classification problems to which traditional statistical methods have long been applied. In building neural network model, the selection of independent and dependent variables should be approached with great care and should be treated as a model construction process. Irrespective of the efficiency of a learning procedure in terms of convergence, generalization and stability, the ultimate performance of the estimator will depend on the relevance of the selected input variables and the quality of the data used. Approaches developed in statistical methods such as correlation analysis and stepwise selection method are often very useful. These methods, however, may not be the optimal ones for the development of neural network models. In this paper, we propose a genetic algorithms approach to find an optimal or near optimal input variables for neural network modeling. The proposed approach is demonstrated by applications to bankruptcy prediction modeling. Our experimental results show that this approach increases overall classification accuracy rate significantly.

• Journal of KIISE:Computing Practices and Letters
• /
• v.14 no.2
• /
• pp.246-250
• /
• 2008

This paper presents the design of network vulnerability analysis system which can integrate various vulnerability assessment tools to improve the preciseness of the vulnerability scan result. Manual checking method performed by a security expert is the most precise and safe way. But this is not appropriate for the large-scale network which has a lot of systems and network devices. Therefore automatic scanning tool is recommended for fast and convenient use. The scanning targets may be different according to the kind of vulnerability scanners, or otherwise even for the same scanning target, the scanning items and the scanning results may be different by each vulnerability scanner, Accordingly, there are the cases in which various scanners, instead of a single scanner, are simultaneously utilized with the purpose of complementing each other. However, in the case of simultaneously utilizing various scanners on the large-scale network, the integrative analysis and relevance analysis on vulnerability information by a security manager becomes time-consumable or impossible. The network vulnerability analysis system suggested in this paper provides interface which allows various vulnerability assessment tools to easily be integrated, common policy which can be applied for various tools at the same time, and automated integrative process.

• The Journal of the Acoustical Society of Korea
• /
• v.20 no.4
• /
• pp.48-53
• /
• 2001

In this paper, selectively attentive learning method has been proposed to improve the learning speed of multilayer Perceptron based on the error backpropagation algorithm. Three attention criterions are introduced to effectively determine which set of input patterns is or which portion of network is attended to for effective learning. Such criterions are based on the mean square error function of the output layer and class-selective relevance of the hidden nodes. The acceleration of learning time is achieved by lowering the computational cost per iteration. Effectiveness of the proposed method is demonstrated in a speaker adaptation task of isolated word recognition system. The experimental results show that the proposed selective attention technique can reduce the learning time more than 60% in an average sense.

• Journal of Digital Convergence
• /
• v.11 no.5
• /
• pp.313-318
• /
• 2013

QoS in wireless multimedia sensor networks is an important issue to enhance streaming media service. This paper proposes a new 2-layer based QoS evaluation scheme for enhancing the streaming media QoS of wireless multimedia sensor networks. The proposed scheme performs the fuzzy relevance to control the streaming between application layer and network layer, and it performs 2-layer mapping process to enhance the transmission reliability and throughput. The simulation results show that the proposed scheme achieves improved performance in packet control ratio, transmission reliability, and delay overhead ratio compared with those of other existing schemes.

• Journal of Advanced Navigation Technology
• /
• v.15 no.3
• /
• pp.364-371
• /
• 2011

In P2P distribution mobile networks, QoS of streaming media services are under heavy influence from overheads such as congestion, latency, and interference. The problem is further complicated by the fact that the popularity of media objects changes over time. This paper proposes a new FSMS+ (Fuzzy Similarity-based Multi-level Streaming Scheme) which minimizes performance degradation of streaming services due to overhead. We then utilize fuzzy similarity-based relevance that can dynamically stream the streaming media object with minimum overhead. The simulation result showed that the proposed scheme has better performance in retransmission rate, congestion control rate and latency rate than the other existing methods of distance method, DC (disk caching) method, and prefix method.

• Journal of Korean Society of Water and Wastewater
• /
• v.33 no.2
• /
• pp.159-167
• /
• 2019

In this paper the potential of the principal component analysis(PCA) technique for the application of detecting leaks in water pipe networks was evaluated. For this purpose the PCA was conducted to evaluate the relevance of the calculated outliers of a PCA model utilizing the recorded pipe flows and the recorded pipe leak incidents of a case study water distribution system. The PCA technique was enhanced by applying the computational algorithms developed in this study which were designed to extract a partial set of flow data from the original 24 hour flow data so that the effective outlier detection rate was maximized. The relevance of the calculated outliers of a PCA model and the recorded pipe leak incidents was analyzed. The developed algorithm may be applied in determining further leak detection field work for water distribution blocks that have more than 70% of the effective outlier detection rate. However, the analysis suggested that further development on the algorithm is needed to enhance the applicability of the PCA in detecting leaks by considering series of leak reports happening in a relatively short period.

• KIPS Transactions on Software and Data Engineering
• /
• v.2 no.5
• /
• pp.341-346
• /
• 2013

In this paper, we propose a rule extraction method using a modified Fuzzy Min-Max (FMM) neural network. The suggested method supplements the hyperbox definition with a frequency factor of feature values in the learning data set. We have defined a relevance factor between features and pattern classes. The proposed model can solve the ambiguity problem without using the overlapping test process and the contraction process. The hyperbox membership function based on the fuzzy partitions is defined for each dimension of a pattern class. The weight values are trained by the feature range and the frequency of feature values. The excitatory features and the inhibitory features can be classified by the proposed method and they can be used for the rule generation process. From the experiments of sign language recognition, the proposed method is evaluated empirically.

• Proceedings of the Korean Society for Emotion and Sensibility Conference
• /
• 2000.04a
• /
• pp.240-243
• /
• 2000

An apparel products retrieval system was proposed in which users can refer to products using Kansei evaluation values. The system adopts relevance feedback using history of the retrieval to learn the tendency of user evaluation. The system is based on a vector space retrieval model using products images expression as semantic scales. The system makes a query from user inputting information and retrieves closest products from the database. Revising algorithms of the difference method. linear multiple regression performed to investigate the effectiveness and criteria of the search. As a result of evaluation of the accuracy, it was found that the linear multiple regression and the neural network models are effective for the retrieval considering the individual Kansei.

• Ocean Systems Engineering
• /
• v.5 no.1
• /
• pp.41-54
• /
• 2015

This study presents the development of predictive models for uplift capacity of suction caisson in clay using an artificial intelligence technique, extreme learning machine (ELM). Other artificial intelligence models like artificial neural network (ANN), support vector machine (SVM), relevance vector machine (RVM) models are also developed to compare the ELM model with above models and available numerical models in terms of different statistical criteria. A ranking system is presented to evaluate present models in identifying the 'best' model. Sensitivity analyses are made to identify important inputs contributing to the developed models.

• Proceedings of the Korean Information Science Society Conference
• /
• 2004.10a
• /
• pp.235-237
• /
• 2004

Microarray 로 표현되는 유전자 발현 데이터는 일반적으로 샘플(sample) 수에 비해 많은 수의 유전자를 포함한다. 피처 추출은 이러한 데이터에 기계학습 방법론을 효과적으로 적용하기 위한 방법 중 하나로, 학습성능을 향상시키고 계산 시간을 줄일 수 있을 뿐만 아니라 중요한 피처들을 발견할 수 있다는 점에서 큰 의미를 갖는다. 본 연구에서는 베이지안 신경망(Bayesian Neural Network)에 기반 한 자동유효성탐지(Automatic Relevance Detection, ARD) 기법을 사용하여 유전자 발현 데이터에서 학습 오류를 줄이는 동시에 학습에 필요한 최소한의 유전자 집합을 추출할 수 있는 방법을 제시했다. CAMDA 2003에서 제시된 폐종양 환자의 유전자 발현 데이터에 대해 실험한 결과, 12600 개의 유전자 중에서 가장 중요하다고 여겨지는 187 개의 유전자를 발견했으며, 높은 학습성능을 달성했다.