• 한국멀티미디어학회논문지
• /
• 제25권6호
• /
• pp.886-893
• /
• 2022

Outlier detection refers to the task of detecting data that deviate significantly from the normal data distribution. Most outlier detection methods compute an outlier score which indicates the degree to which a data sample deviates from normal. However, setting a threshold for an outlier score to determine if a data sample is outlier or normal is not trivial. In this paper, we propose a binary prediction method for outlier detection based on spectral clustering and one-class SVM ensemble. Given training data consisting of normal data samples, a clustering method is performed to find clusters in the training data, and the ensemble of one-class SVM models trained on each cluster finds the boundaries of the normal data. We show how to obtain a threshold for transforming outlier scores computed from the ensemble of one-class SVM models into binary predictive values. Experimental results with high dimensional text data show that the proposed method can be effectively applied to high dimensional data, especially when the normal training data consists of different shapes and densities of clusters.

• 대한치과의사협회지
• /
• 제19권1호통권140호
• /
• pp.67-71
• /
• 1981

This study was undertaken to find out the characteristic craniofacial morphology of Class II Division 2 malocclusions in children by means of roentgenocephalometry. The subjects consisted of twelve boys and thirteen girls with Class Ii, Division 2, thirty seven boys and fifty three girls with Class II, Division 1, and forty six boys and eighty one girls with normal occlusion, ranged from 10 years old to 18 years old. The following results were obtained; 1. The anteroposterior relationship of the maxilla to the cranium in the Class II, Division 2 malocclusion similar to the normal occlusion, but the mandible was the posterior position in th the cranial anatomy. 2. There were no significant differences in the anteroposterior relationship of the maxilla and the mandible between Class II, Division 1 and Class II, Division 2 malocclusions. 3. In Class II, Division 1 the axes of maxillary incisors showed labial inclination, but lingual inclination in Class II, Division 2 malocclusions. 4. Overbite was prominent one in Class II, Division 2, on the other hand overjet was distinguished in Class II, Division 1.

• 대한수학회보
• /
• 제60권2호
• /
• pp.315-323
• /
• 2023

First, we recall the results for prime-producing polynomials related to class number one problem of quadratic fields. Next, we give the relation between prime-producing cubic polynomials and class number one problem of the simplest cubic fields and then present the conjecture for the relations. Finally, we numerically compare the ratios producing prime values for several polynomials in some interval.

• East Asian mathematical journal
• /
• 제37권5호
• /
• pp.613-617
• /
• 2021

Abstract. For a positive square-free integer m, let K = ℚ($\sqrt{m}$) be a real quadratic field. The relative class number H_d(f) of K of discriminant d is the ratio of class numbers 𝒪_K and 𝒪_f, where 𝒪_K is the ring of integers of K and 𝒪_f is the order of conductor f given by ℤ + f𝒪_K. In 1856, Dirichlet showed that for certain m there exists an infinite number of f such that the relative class number H_d(f) is one. But it remained open as to whether there exists such an f for each m. In this paper, we give a result for existence of real quadratic field ℚ($\sqrt{m}$) with relative class number one where the period of continued fraction expansion of $\sqrt{m}$ is 6.

• 대한치과교정학회지
• /
• 제9권1호
• /
• pp.85-98
• /
• 1979

This investigation was designed to compare the craniofacial and dental morphology of class III malocclusion with that of normal occlusin in children, and to determine the incidence of various class III craniofacial skeletal patterns. The material selected for this study consisted in standard lateral cephalograms of eighty two Korean children, forty one boys and forty one girls, aged 10 through 12 years, having class III malocclusion, and forty two Korean children, twenty boys and twenty two girls, with normal occlusion in the same age. Using the tracings of the standard lateral cephalograms, various angular and linear measurements were recorded, tabulated and statistically analyzed, and then the class III craniofacial skeletal morphology was divided into various patterns by the degree of SNA and SNB, which respectively were below, within or beyond the normal range of those of normal occlusion. The following characteristics of the craniofacial and dental morphology of class III malocclusion were observed. 1. The cranial base length of class III malocclusion was smaller than that of normal occlusion, and the small saddle angle was a characteristic figure of class III malocclucion. 2. Maxillary length of class III malocclusion was smaller than that of normal occlusion, and point A was retropositioned relative to cranial base but not PNS in class III malocclusion. Maxillary base inclination was not significantly different between the two, but occlusal plane to palatal plane was small in class III malocciusion. 3. The mandibular body length shown no difference between the two, but the mandibular body positioned anteriorly relative to cranial base in class III malocclusion. Ramus height, gonial angle, and mandibular effective length were large in class III malocclusion. Mandibular plane angle and joint angle had no difference between the two, and occlusal plane to mandibular plane angle was large in class III malocclusion. 4. Maxillary incisor inclination was not significantly different between class III malocclusion and normal occlusion, but mandibular incisors positioned and inclined lingually and consequently interincisal angle was large in class III malocclusion. 5. Class III malocclusion was divided into six categories of craniofacial skeletal pattern. The most common class III pattern was found to be one in which the maxilla was within the normal range of prognathism while the mandible extended beyond this range. The pattern in which the maxilla was below the normal range of prognathism while the mandible was within this range was approximately one fifth of the class III sample.

• 한국융합학회논문지
• /
• 제9권6호
• /
• pp.33-38
• /
• 2018

본 논문에서는 단일 클래스만을 학습하여 차량에 대한 새로운 공격을 탐지한다. 분류 성능 평가를 위해 Car-Hacking 데이터셋을 사용한다. Car-Hacking 데이터셋은 실제 차량의 OBD-II 포트를 통해 CAN (Controller Area Network) 트래픽을 로깅하여 생성된다. 이 데이터셋에는 네 가지 공격 유형이 포함된다. 실험에 사용한 단일 클래스 분류기법은 정상 클래스만을 학습하여 비정상인 공격 클래스를 분류해내는 비지도 학습이다. 비지도 학습 방법을 사용하는 경우에 훈련 과정에서 네거티브 인스턴스를 사용하지 않기 때문에 고효율의 분류 성능을 내는 것은 어렵다. 하지만, 비지도 학습은 라벨이 없는 새로운 공격 데이터를 분류하는데 적합한 장점이 있다. 본 연구에서는 네트워크 침입탐지 시스템에서 서명기반의 규칙으로 탐지하기 어려운 새로운 공격 유형을 탐지하기 위해 단일 클래스 분류기를 사용한다. 제안 방법은 새로운 공격을 모두 탐지하고 정상데이터에 대해서도 효율적인 분류 성능을 보이는 파라미터 조합을 제시한다.

• 한국융합학회논문지
• /
• 제9권11호
• /
• pp.15-21
• /
• 2018

본 논문에서는 단일 클래스만을 학습하여 네트워크 침입탐지 시스템 상에서 새로운 비정상 행위를 탐지하는 것을 목표로 한다. 분류 성능 평가를 위해 KDD CUP 1999 데이터셋을 사용한다. 단일 클래스 분류는 정상 클래스만을 학습하여 공격 클래스를 분류해내는 비지도 학습 방법 중 하나이다. 비지도 학습의 경우에는 학습에 네거티브 인스턴스를 사용하지 않기 때문에 상대적으로 높은 분류 효율을 내는 것이 어렵다. 하지만, 비지도 학습은 라벨이 없는 데이터를 분류하는데 적합한 장점이 있다. 본 연구에서는 서포트벡터머신 기반의 단일 클래스 분류기와 밀도 추정 기반의 단일 클래스 분류기를 사용한 실험을 통해 기존에 없던 새로운 공격에 대한 탐지를 한다. 밀도 추정 기반의 분류기를 사용한 실험이 상대적으로 더 좋은 성능을 보였고, 신규 공격에 대해 낮은 FPR을 유지하면서도 약 96%의 탐지율을 보인다.

• 한국통신학회논문지
• /
• 제21권3호
• /
• pp.629-642
• /
• 1996

ATM switch handles the traffic for a wide range of appliations with different QOS(Quality-of-Service) requirements. In ATM switch, the priority control mechanism is needed to improve effectively the required QOS requirements. In this paper, we propose a priority control mechanism using the cell transfer ratio type and discard threshold in order to archive the cell loss probability requirement and the delay requirement of each service class. The service classes of our concern are the service class with high time priority(class 1) and the service class with high loss priority control mechanism, cells for two kind of service classes are stored and processed within one buffer. In case cells are stored in the buffer, cells for class 2 are allocated in the stored and processed within one buffer. In case cells are stored in the buffer, cells for class 2 are allocated in the shole range of the buffer and cells for class 1 are allocated up to discard threshold of the buffer. In case cells in the buffer are transmitted, one cell for class 1 is transmitted whenever the maximum K cells for class 2 are transmitted consecutively. We analyze the time delay and the loss probability for each class of traffic using Markov chain. The results show that the characteristics of the mean cell delay about cells for class 1 becomes better and that of the cell loss probability about cells for class 2 becomes better by selecting properly discard threshold of the buffer and the cell transfer ratio according to the condition of input traffic.

• 대한치과교정학회지
• /
• 제11권2호
• /
• pp.101-108
• /
• 1981

This investigation was designed to compare the calcification degree of maxillary second permanent molar to mandibular second permanent molar in skeletal Class III Malocclusion. The material selected for this study consisted in standand lateral cephalogram study model and orthopantomogram of two hundred fifty seven Korean Children, one hundred twenty one boys and one hundred twenty four girls, aged 6 through 12 years, having skeletal Class III Malocclusion. On the basis of findigs of this study, the following results were obtained 1. In the stage of completion of crown, there was no significant difference in calcification degree between maxillary second molar and mandibular second molar of both boys and girls in skeletal Class III Malocclusion. 2. From 8 years of age at the stage of beginning root formation to 12 years of age, the calcification degree of mandibular second molar was more advanced than Maxillary second molar of both boys and girls in skeletal Class III Malocclusion.

• 대한산업공학회지
• /
• 제16권1호
• /
• pp.99-106
• /
• 1990

With two-class-based storage assignment policy and dual command cycle in Automated Storage/Retrieval Systems(AS/RS), the problem of determining the region dedicated for class-one item is considered. First, the expected travel time of the S/R machine is derived when the boundary of the class-one region is square. Secondly, a heuristic procedure is proposed which determines sequentially the class-one region in a discrete rack. An application of the procedure generates leaf shape region which confirms that the L-shape partition is not necessarily optimal.