• 한국정보처리학회:학술대회논문집
• /
• 한국정보처리학회 2023년도 춘계학술발표대회
• /
• pp.2-4
• /
• 2023

Sequence data plays an important role in the field of intelligence, especially for industrial control, traffic control and other aspects. Finding abnormal parts in sequence data has long been an application field of AI technology. In this paper, we propose an anomaly detection method for sequence data using a diffusion model. The diffusion model has two major advantages: interpretability derived from rigorous mathematical derivation and unrestricted selection of backbone models. This method uses the diffusion model to predict and reconstruct the sequence data, and then detects the abnormal part by comparing with the real data. This paper successfully verifies the feasibility of the diffusion model in the field of anomaly detection. We use the combination of MLP and diffusion model to generate data and compare the generated data with real data to detect anomalous points.

• Journal of the Korean Data and Information Science Society
• /
• 제23권2호
• /
• pp.385-392
• /
• 2012

Logistic regression is a well known binary classification method in the field of statistical learning. Mixed-effect regression models are widely used for the analysis of correlated data such as those found in longitudinal studies. We consider kernel extensions with semiparametric fixed effects and parametric random effects for the logistic regression. The estimation is performed through the penalized likelihood method based on kernel trick, and our focus is on the efficient computation and the effective hyperparameter selection. For the selection of optimal hyperparameters, cross-validation techniques are employed. Numerical results are then presented to indicate the performance of the proposed procedure.

• 한국국방경영분석학회지
• /
• 제25권1호
• /
• pp.75-86
• /
• 1999

The selection criteria of an Authorized Stockage List (ASL) in the Army is based on Army Regulation(AR)409. However, the current selection method of ASL is not considered in cost, weight and volume of repair parts. This paper is focused on developing for a new selection model taking account of cost, weight and volume of repair parts. Goal programming is utilized in order to consider weighted priorities. Different units of cost, and volume are normalized for using weighing value. Real data of a field division are applied to the model. Results of the new selection model are more reduced in cost, weight and volume than those of the previous method.

• 한국수학사학회지
• /
• 제18권2호
• /
• pp.95-106
• /
• 2005

본 논문에서는 수학을 기반으로 한 데이터베이스의 지식탐사 절차를 통하여 데이터의 선택, 정제, 통합, 변환, 축소, 데이터 마이닝 기법의 선택과 적용 및 모형의 평가에 관한 개념과 방법론을 소개하고 수학의 한 분야로서 통계학의 역할과 적용방법에 관하여 연구하고자 한다. 또한 오늘날 관심이 대상이 되고 있는 데이터 마이닝의 역사와 수학적 배경, 통계 및 정보 기술을 이용한 데이터 마이닝의 주요 모델링 기법, 실용적 응용 분야 및 적용 사례 그리고 데이터 마이닝과 통계의 차이점에 관하여 조사하고 논하고자 한다.

• Journal of Information Technology Applications and Management
• /
• 제28권6호
• /
• pp.97-116
• /
• 2021

Due to the decrease in the school-age population and government regulations, universities have made great efforts to secure their own competitiveness. In particular, the selection of universities with financial support based on the recent evaluation of the Ministry of Education has become a major concern enough to affect the existence of the university itself. This paper extracts three-year data from 124 major private universities nationwide, and quantitatively analyzes the variables of major universities selected as self-improvement universities, competency reinforcement universities, and universities with limited financial support. As a result of estimating the selection of self-powered universities using the ordered logit model by hierarchically inputting 12 variables, student competitiveness in the metropolitan area (1.318^**), Educational Restitution Rate (4.078^***), University operation expenditure index rate (1.088^***) values were found. Significant positive coefficient values were found in the admission enrollment rate (45.98^***) and the enrollment rate (13.25^***). As a result of analyzing the marginal effects, the increase in the rate of reduction of education costs has always been positive in the selection of self-powered universities, but it was observed that the rate of increase decreases in areas of increase of 150% or more. On the contrary, the probability of becoming a Em-powered university was negative in all sectors, but on the contrary, it was analyzed that marginal effects increased at the same time point. On the other hand, the employment rate of graduates was not able to find direct significance with the result of the selection of Self powered universities. Through this paper, it is expected that each university will analyze the possibility and shortcomings of the selection of Self powered universities in policy making, and in particular, the risk of dropout of selection for the vulnerable field can be predicted using marginal effects. It can be used as major research data for both university evaluators, university officials and students.

• Asian-Australasian Journal of Animal Sciences
• /
• 제10권6호
• /
• pp.567-570
• /
• 1997

A multiple trait animal model having milk yield and lactation length was used to estimate genetic parameters using data from four institutional herds and four field recording centers. Response to selection for milk yield alone and in combination with lactation length was estimated by using principles of genetic theory. Lactation records (n = 2,353) adjusted for age at calving to 60 months were utilized. Milk yield was 17% heritable with repeatability of 0.44. Lactation length had a low heritability of 0.06 with repeatability of 0.16. Genetic correlation between the two traits was 0.70. Selection response in milk yield can be improved slightly (103.8 vs 102.8 kg) when information on covariance with lactation length is used together with the information on milk yield.

• 한국지반공학회:학술대회논문집
• /
• 한국지반공학회 2009년도 세계 도시지반공학 심포지엄
• /
• pp.27-50
• /
• 2009

This paper describes an adaptive management approach for predicting, monitoring, and controlling ground movements associated with excavations in urban areas. Successful use of monitoring data to update performance predictions of supported excavations depends equally on reasonable numerical simulations of performance, the type of monitoring data used as observations, and the optimization techniques used to minimize the difference between predictions and observed performance. This paper summarizes each of these factors and emphasizes their inter-dependence. Numerical considerations are described, including the initial stress and boundary conditions, the importance of reasonable representation of the construction process, and factors affecting the selection of the constitutive model. Monitoring data that can be used in conjunction with current numerical capabilities are discussed, including laser scanning and webcams for developing an accurate record of construction activities, and automated and remote instrumentations to measure movements. Self-updating numerical models that have been successfully used to compute anticipated ground movements, update predictions of field observations and to learn from field observations are summarized. Applications of these techniques from case studies are presented to illustrate the capabilities of this approach.

• 천문학회보
• /
• 제41권2호
• /
• pp.64.2-64.2
• /
• 2016

Optical variability is one way to probe the nature of the central engine of AGN at smaller linear scales and previous studies have shown that optical variability is more prevalent at longer timescales and at shorter wavelengths. Especially, intra-night variability can be explained through the damped random walk model but small samples and inhomogeneous data have made constraining this model hard. To understand the properties and physical mechanism of optical variability, we are performing the KMTNet Active Nuclei Variability Survey (KANVaS). Test data of KMTNet in the COSMOS field was obtained over 2 separate nights during 2015, in B, V, R, and I bands. Each night was composed of 5 and 9 epochs with ~30 min cadence. To find AGN in the COSMOS field, we applied multi-wavelength selection methods. Different selection methods means we are looking different region in unification model of AGN, and 100~120, 400~500, 50~100 number of AGN are detected in X-ray, mid-infrared, and radio selection of AGN, respectively. We performed image convolution to reflect seeing fluctuation, then differential photometry between the selected AGN and nearby stars to achieve photometric uncertainty ~0.01mag. We employed one of the standard time-series analysis tools to identify variable AGN, chi-square test. Preliminarily results indicate that intra-night variability is found for X-ray selected, Type1 AGN are 23.6%, 26.4%, 21.3% and 20.7% in the B, V, R, and I band, respectively. The majority of the identified variable AGN are classified as Type 1 AGN, with only a handful of Type 2 AGN showing evidence for variability. The work done so far confirms that there are type and wavelength dependence of intra-night optical variability of AGN.

• 한국산업정보학회논문지
• /
• 제21권6호
• /
• pp.13-21
• /
• 2016

특징선택은 데이터 마이닝, 기계학습 분야에서 가장 중요한 이슈 중 하나로, 원본 데이터에서 가장 좋은 분류 성능을 보여줄 수 있는 특징들을 찾아내는 방법이다. 본 논문에서는 정보 입자성을 기반으로 한 neighborhood 러프집합 모델을 이용한 특징선택 방법을 제안한다. 제안된 방법의 효과성은 5,252명의 유방 초음파 영상으로부터 추출된 298가지의 특징들 중에서 유방 종양의 진단과 관련된 유용한 특징들을 선택하는 문제에 적용되었다. 실험결과 19가지의 진단적 특징을 찾을 수 있었고, 이때에 평균 분류 정확성은 97.6%를 보였다.

• 국제학술발표논문집
• /
• The 3th International Conference on Construction Engineering and Project Management
• /
• pp.153-160
• /
• 2009

The demand for super tall building construction is increasing worldwide. There has been a constant request for achieving early payback on investment by shortening the construction time. This pertains especially for the case of huge investment projects such as super tall building construction. It is very important to shorten the construction time for the building framework, which requires substantial construction time and cost, and this is directly related to the establishment of an optimum lift plan for construction. When there is a problem in the selection of the lift equipment, it is almost impossible to revise the selection, resulting in a possible failure of the project. Thus, the purpose of this study is to analyze the function and logic for the development of the process for the selection of lift equipment for super tall building projects and further development of making the analyzed process into a system. In line with this research objective, the process of selecting the optimum lift equipment by domestic construction company was investigated and analyzed as well as collecting the actual field data. The actual data were obtained by sensors installed on tower cranes at three construction sites with the help from the construction company.