• 대한원격탐사학회지
• /
• 제20권5호
• /
• pp.353-359
• /
• 2004

This study proposes an approach to unsupervisedly estimate the number of classes and the parameters of defining the classes in order to train the classifier. In the proposed method, the image is segmented using a spatial region growing based on hierarchical clustering, and fuzzy training is then employed to find the sample classes that well represent the ground truth. For cluster validation, this approach iteratively estimates the class-parameters in the fuzzy training for the sample classes and continuously computes the log-likelihood ratio of two consecutive class-numbers. The maximum ratio rule is applied to determine the optimal number of classes. The experimental results show that the new scheme proposed in this study could be used to select the regions with different characteristics existed on the scene of observed image as an alternative of field survey that is so expensive.

• 대한수학회보
• /
• 제56권4호
• /
• pp.815-827
• /
• 2019

It is well known that the denominator of the Dedekind sum s(c, d) divides 2 gcd(d, 3)d and that no smaller denominator independent of c can be expected. In contrast, here we prove that we usually get a smaller denominator in S(H, d), the sum of the s(c, d)'s over all the c's in a subgroup H of order n > 1 in the multiplicative group $(\mathbb{Z}/d\mathbb{Z})^*$. First, we prove that for p > 3 a prime, the sum 2S(H, p) is a rational integer of the same parity as (p-1)/2. We give an application of this result to upper bounds on relative class numbers of imaginary abelian number fields of prime conductor. Finally, we give a general result on the denominator of S(H, d) for non necessarily prime d's. We show that its denominator is a divisor of some explicit divisor of 2d gcd(d, 3).

• 전기학회논문지
• /
• 제68권4호
• /
• pp.523-531
• /
• 2019

This paper deals with a wound-field synchronous machines(WFSM), with an electromagnet on its salient rotor, as an alternative to a permanent magnet in the rotor. We then examine the power performance characteristics, loss characteristics, V-curves and large short-circuit ratios for a large-scale synchronous generator, considering the leading and lagging operations, based on the finite-element method. We predict the performance of a 100MVA-class generator based on the operating range for a constant short-circuit ratio. At the last, We compared with the electromagnetic characteristics of three model according to number of slots.

• 대한수학회보
• /
• 제61권3호
• /
• pp.611-619
• /
• 2024

For a finite group G, let 𝜓(G) denote the sum of element orders of G. If ${\psi}^{{\prime}{\prime}}(G)\,=\,{\frac{\psi(G)}{{\mid}G{\mid}^2}}$, we show here that the image of 𝜓'' on the class of all Dihedral groups whose order is twice a composite number greater than 4 is dense in $[0,\,{\frac{1}{4}}]$. We also derive some properties of 𝜓'' on the class of all dihedral groups whose order is twice a prime number.

• 대한치과교정학회지
• /
• 제22권3호
• /
• pp.627-646
• /
• 1992

The purpose of this study was to investigate the relationship among the activity of the craniofacial muscle and craniofacial form and occlusal state. In this study, subjects were consisted of 23 male adults with skeletal Class III malocclusion and 30 male adults with normal occlusion. The measurements in oral exam, lateral ceghalogram, and E.M.G. recordings of anterior temporal, masseter, and upper lip muscles at rest position, clenching in centric occlusion, chewing of gum, swallowing of juice, were analyzed with SPSS system. The results were as follows: 1. At rest position upper lip muscle activity of skeletal Class III group was significantly higher than that of normal group. 2. Both clenching and chewing masseter and temporal muscle activity of normal group were significantly higher than that of skeletal Class III group. 3. During swallowing of juice, upper lip muscle activity of skeletal Class III group were significantly higher than that of normal group. 4. The activities of masseter and anterior temporal muscle during clenching and chewing were significantly correlated with hypodivergent facial form and number of occluded teeth. 5. The activity of upper lip during swallowing had positive correlation with mandibular prognathism.

• 한국지능정보시스템학회:학술대회논문집
• /
• 한국지능정보시스템학회 2007년도 한국지능정보시스템학회
• /
• pp.309-320
• /
• 2007

The data imbalance problem which can be uncounted in data mining classification problems typically means that there are more or less instances in a class than those in other classes. It causes low prediction accuracy of the minority class because classifiers tend to assign instances to major classes and ignore the minor class to reduce overall misclassification rate. In order to solve the data imbalance problem, there has been proposed a number of techniques based on resampling with replacement, adjusting decision thresholds, and adjusting the cost of the different classes. In this paper, we study the feasibility of the combination usage of the techniques previously proposed to deal with the data imbalance problem, and suggest a combination method using genetic algorithm to find the optimal combination ratio of the techniques. To improve the prediction accuracy of a minority class, we determine the combination ratio based on the F-value of the minority class as the fitness function of genetic algorithm. To compare the performance with those of single techniques and the matrix-style combination of random percentage, we performed experiments using four public datasets which has been generally used to compare the performance of methods for the data imbalance problem. From the results of experiments, we can find the usefulness of the proposed method.

• 한국경영과학회지
• /
• 제38권1호
• /
• pp.139-152
• /
• 2013

The purpose of this study is to claim the validity of tuning the architecture of neural network models for multi-class classification. A neural network model for multi-class classification is basically constructed by building a series of neural network models for binary classification. Building a neural network model, we are required to set the values of parameters such as number of hidden nodes and weight decay parameter in advance, which draws special attention as the performance of the model can be quite different by the values of the parameters. For better performance of the model, it is absolutely necessary to have a prior process of tuning the parameters every time the neural network model is built. Nonetheless, previous studies have not mentioned the necessity of the tuning process or proved its validity. In this study, we claim that we should tune the parameters every time we build the neural network model for multi-class classification. Through empirical analysis using wine data, we show that the performance of the model with the tuned parameters is superior to those of untuned models.

• 한국수학교육학회지시리즈D:수학교육연구
• /
• 제5권2호
• /
• pp.87-98
• /
• 2001

The Center for Science Gifted Education (CSGE) of Chongju National University of Education was established in 1998 with the financial support of the Korea. Science & Engineering Foundation (KOSEF). In fact, we had prepared mathematics and science gifted education program beginning in 1997. It was possible due to the commitment of faculty members with an interest in gifted education. Now we have 5 classes in Mathematics, two of which are fundamental, one of which is a strengthened second-grade class gifted elementary school students, and one a fundamental class, and one a strengthened class for gifted middle school students in Chungbuk province. Each class consists of 16 students selected by a rigorous examination and filtering process. Also we have a mentoring system for particularly gifted students in mathematics. We have a number of programs for Super-Saturday, Summer School, Winter School, and Mathematics and Science Gifted Camp. Each program is suitable for 90 or 180 minutes of class time. The types of tasks developed can be divided into experimental, group discussion, open-ended problem solving, and exposition and problem solving tasks. Levels of the tasks developed for talented elementary students in mathematics can be further divided into grade 5 and under, grade 6, and grade 7 and over. Types of the tasks developed can be divided into experimental, group discussion, open-ended problem solving, and exposition and problem solving task. Also levels of the tasks developed for talented elementary students in mathematics can be divided into the level of lower than grade 5, level of grade 6, and level of more than grade 7. Three tasks developed and practiced are reported in this article.

• 복식
• /
• 제22권
• /
• pp.297-312
• /
• 1994

Today though the word ginyoe(gisaeng) was remained as a historic relic but they were firmly existed about 40 years ago and ginyoe's number was about three million in Chosun it's almost near 0.5% of the total population of Chosun. To think that point the ginyoe's so-ciety was considered a special one in the his-tory of Korean woman. The ginyoe as a special social class were specialize in technical art such as dancing and prostitute. Although they were low class they were luxurious slaves whoses clothing was almost equal to that of royal family. They were the leaders of fashion in woman clothing that's because their role was entertainer, This study can be summarized as follows. First ginyoe and public woman's clothing codes were nearly same in koryo but tatally separated in chosun. I could find that was came from those day's moral sprit. Second ginyoe's clothing was not only have luxuriance like royal family but also have unique clothing codes for them. Though they are low class ginyoe could use upper class's clothing codes. But upper class women could'nt use ginyoe's clothing codes are vary various and have their own clothing codes. Third I analyzed ginyoe's clothing codes and then derived 6 ginyoe's clothing messages. They were symbol of wealth symbol of power symbol of occupational function symbol of sexual attraction symbol of social position symbol of freedom.

• 대한수학교육학회지:학교수학
• /
• 제2권2호
• /
• pp.475-487
• /
• 2000

As a result of carrying out t-test against the learning activity that used the learning materials designed by level, significant disparity was found. Thus, it was found that hte effect of this learning activity became more visible with lapse of time. But the major cause by which there appeared small significant disparity as a result of testing is that the units of experiment were limited and that the period of using the learning materials was not long. In an analysis on the results of interest test, the experimental class also appeared to show the average score that was higher than that of the comparative class by 0.10 after converting a decimal point. The outcome of attitude test was that the experimental class showed a higher average score by 0.11, as a result of converting a decimal point, than that of the comparative class. So, a large number of students showed an improved reaction. But, there seemed some problems of the learning materials or the method of progressing the activity in changing under achievement students or students who avoided studying math. When the effect of the level-based learning activity was investigated, more than 80% of the experimental class’s students showed a positive reaction. Thus, it could be judged that students, who felt some burden at studying math, might be served more largely, not by teacher’s uniform instruction, but by an individual learning using the level-based learning materials that enabled them to do a systematic self-learning for themselves.