• The Transactions of The Korean Institute of Electrical Engineers
• /
• v.56 no.12
• /
• pp.2084-2090
• /
• 2007

The DGA (Dissolved Gases Analysis) technique has been widely using for fault diagnosis of the power transformers. Some electric power utility company establishes the criteria of DGA to improve reliability, because of difference of operation environment and design of power transformer. In this paper, we introduce intelligent diagnosis system for DGA result of KEPCO (Korea Electric Power Cooperation). This system can classify patterns type of gases ratio that frequently occurs in recent result of gases analysis using Fuzzy Inference. The classification of Patterns let us know that major causes of gases generation based on type of patterns. Finally, Neural Network based on patterns diagnose transformer. NN was trained using result data of DGA of actually faulted transformers recently. Result of intelligent diagnosis system is right well in comparison with actual inner inspection of transformers.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.17 no.6
• /
• pp.799-803
• /
• 2007

In this paper, we propose a new fuzzy membership function for FSVM(Fuzzy Support Vector Machines). We apply a fuzzy membership to each input point of SVM and reformulate SVM into fuzzy SVM (FSVM) such that different input points can make different contributions to the learning of decision surface. The proposed method enhances the SVM in reducing the effect of outliers and noises in data points. This paper compares classification and estimated performance of SVM, FSVM(1), and FSVM(2) model that are getting into the spotlight in time series prediction.

• Proceedings of the Korean Institute of Intelligent Systems Conference
• /
• 2003.09a
• /
• pp.376-379
• /
• 2003

In this paper we describe DCClass, a tool for fuzzy information granulation with transparency constraints. The tool is particularly suited to solve fuzzy classification problems, since it is able to automatically extract information granules with class labels. For transparency pursuits, the resulting information granules are represented in form of fuzzy Cartesian product of one-dimensional fuzzy sets. As a key feature, the proposed tool is capable to self-determining the optimal granularity level of each one-dimensional fuzzy set by exploiting class information. The resulting fun information granules can be directly translated in human-comprehensible fuzzy rules to be used for class inference. The paper reports preliminary experimental results on a medical diagnosis problem that shows the utility of the proposed tool.

• Proceedings of the Korean Institute of Intelligent Systems Conference
• /
• 2003.09a
• /
• pp.595-598
• /
• 2003

A new classifier is constructed by using a generalized regression neural network (GRNN) in conjunction with a random generator (RC). The RG played a role of generating a number of sets of random spreads given a range for gaussian functions in the pattern layer, The range experimentally varied from 0.4 to 1.4. The DNA sequences consisted 4 types of promoters. The performance of classifier is examined in terms of total classification sensitivity (TCS), and individual classification sensitivity (ICS). for comparisons, another GRNN classifier was constructed and optimized in conventional way. Compared GRNN, the RG-GRNN demonstrated much improved TCS along with better ICS on average.

• International Journal of Fuzzy Logic and Intelligent Systems
• /
• v.4 no.1
• /
• pp.18-22
• /
• 2004

It is difficult to understand a person's emotion, since it is subjective and vague. Therefore, we are proposing a method by which will effectively classify human emotions into two types (that is, single emotion and composition emotion). To verify validity of te proposed method, we conducted two experiments based on general inference and $\alpha$-cut, and compared the experimental results. In the first experiment emotions were classified according to fuzzy inference. On the other hand in the second experiment emotions were classified according to $\alpha$-cut. Our experimental results showed that the classification of emotion based on a- cut was more definite that that based on fuzzy inference.

• International Journal of Fuzzy Logic and Intelligent Systems
• /
• v.16 no.2
• /
• pp.81-86
• /
• 2016

Term weighting is a popular technique that effectively weighs the term features to improve accuracy in document classification. While several successful term weighting algorithms have been suggested, none of them appears to perform well consistently across different data domains. In this paper we propose several reasonable methods to combine different term weight vectors to yield a robust document classifier that performs consistently well on diverse datasets. Specifically we suggest two approaches: i) learning a single weight vector that lies in a convex hull of the base vectors while minimizing the class prediction loss, and ii) a mini-max classifier that aims for robustness of the individual weight vectors by minimizing the loss of the worst-performing strategy among the base vectors. We provide efficient solution methods for these optimization problems. The effectiveness and robustness of the proposed approaches are demonstrated on several benchmark document datasets, significantly outperforming the existing term weighting methods.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.18 no.5
• /
• pp.638-641
• /
• 2008

In this paper, a model to classify the method using the fuzzy TAM with preprocessing of data was developed. The preprocessing method can be divide the problem using the characteristics in the case of category type factor. In case of continuous type factor, if there was exist factor's range which is not overlapping by class, the data belong to the range was fixed and eliminated in classification. After these preprocessing of data, classified operation of Fuzzy TAM is performed.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.11 no.7
• /
• pp.633-640
• /
• 2001

In this paper we examine the machine learning issues raised by the domain of the Intrusion Detection Systems(IDS), which have difficulty successfully classifying intruders. There systems also require a significant amount of computational overhead making it difficult to create robust real-time IDS. Machine learning techniques can reduce the human effort required to build these systems and can improve their performance. Genetic algorithms are used to improve the performance of search problems, while data mining has been used for data analysis. Data Mining is the exploration and analysis of large quantities of data to discover meaningful patterns and rules. Among the tasks for data mining, we concentrate the classification task. Since classification is the basic element of human way of thinking, it is a well-studied problem in a wide variety of application. In this paper, we propose a classifier system based on genetic algorithm, and the proposed system is evaluated by applying it to IDS problem related to classification task in data mining. We report our experiments in using these method on KDD audit data.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.22 no.4
• /
• pp.515-520
• /
• 2012

In brain computer interface (BCI) system, the most important part is classification of human thoughts in order to translate into commands. The more accuracy result in classification the system gets, the more effective BCI system is. To increase the quality of BCI system, we proposed to reduce noise and artifact from the recording data to analyzing data. We used auditory stimuli instead of visual ones to eliminate the eye movement, unwanted visual activation, gaze control. We applied independent component analysis (ICA) algorithm to purify the sources which constructed the raw signals. One of the most famous spatial filter in BCI context is common spatial patterns (CSP), which maximize one class while minimize the other by using covariance matrix. ICA and CSP also do the filter job, as a raw filter and refinement, which increase the classification result of linear discriminant analysis (LDA).

• Journal of the Korean Institute of Intelligent Systems
• /
• v.21 no.4
• /
• pp.487-492
• /
• 2011

In this paper, we proposed a new design methodology to improve the classification performance of the Nearest Prototype Classifier which is one of the simplest classification algorithm. To optimize the position vectors of the prototypes in the nearest prototype classifier, we use the differential evolutionary algorithm. The optimized position vectors of the prototypes result in the improvement of the classification performance. The new method to determine the class labels of the prototypes, which are defined by the differential evolutionary algorithm, is proposed. In addition, the experimental application covers a comparative analysis including several previously commonly encountered methods.