• 대한수학회지
• /
• 제48권1호
• /
• pp.147-158
• /
• 2011

In this paper, we consider a multi-dimensional diffusion process in a self-similar random environment and prove a limit theorem for the shape of the full trajectory of the diffusion by using the localization phenomenon.

• International Journal of Advanced Culture Technology
• /
• 제10권4호
• /
• pp.361-368
• /
• 2022

This study seeks to determine the research trend of late adolescents by utilizing big data. Also, seek for research trends related to activity participation, treatment, and mediation to provide academic implications. For this process, gathered 1.000 academic papers and used TF-IDF analysis method, and the topic modeling based on co-occurrence word network analysis method LDA (Latent Dirichlet Allocation) to analyze. In conclusion this study conducted analysis of activity participation, treatment, and mediation of late adolescents by TF-IDF analysis method, co-occurrence word network analysis method, and topic modeling analysis based on LDA(Latent Dirichlet Allocation). The results were proposed through visualization, and carries significance as this study analyzed activity, treatment, mediation factors of late adolescents, and provides new analysis methods to figure out the basic materials of activity participation trends, treatment, and mediation of late adolescents.

• Communications for Statistical Applications and Methods
• /
• 제29권3호
• /
• pp.287-299
• /
• 2022

In radiation epidemiology, the excess relative risk (ERR) model is used to determine the dose-response relationship. In general, the dose-response relationship for the ERR model is assumed to be linear, linear-quadratic, linear-threshold, quadratic, and so on. However, since none of these functions dominate other functions for expressing the dose-response relationship, a Bayesian semiparametric method using splines has recently been proposed. Thus, we improve the Bayesian semiparametric method for the selection of the tuning parameters for splines as the number and location of knots using a Bayesian knot selection method. Equally spaced knots cannot capture the characteristic of radiation exposed dose distribution which is highly skewed in general. Therefore, we propose a nonparametric Bayesian knot selection method based on a Dirichlet process mixture model. Inference of the spline coefficients after obtaining the number and location of knots is performed in the Bayesian framework. We apply this approach to the life span study cohort data from the radiation effects research foundation in Japan, and the results illustrate that the proposed method provides competitive curve estimates for the dose-response curve and relatively stable credible intervals for the curve.

• 한국ITS학회 논문지
• /
• 제20권4호
• /
• pp.57-70
• /
• 2021

본 연구는 민원데이터를 분석함으로써 교통경찰에 대한 국민의 치안 수요를 탐색하고자 하였다. 이를 위해 교통경찰 관련 국민신문고 민원데이터 2,062건을 대상으로, 토픽모델링 방법 중 하나인 잠재 디리클레 할당(Latent Dirichlet Allocation)을 통해 주요 토픽을 추출하고 높은 비중을 차지한 위반신고에 대해 추가분석을 시도하였다. 이 과정에서 키워드와 대표문서의 일관성과 합치성을 함께 고려하였다. 분석 결과 교통경찰 관련 민원은 시설개선, 신호에 따른 교차로통행방법, 번호판 영치, 개인형 이동장치 등 41개의 토픽으로 분류할 수 있었다. 교차로내 위반과 이륜자동차의 위반에 대한 단속을 강화하고 무인교통단속장비, 횡단보도, 신호등의 설치 및 운영에 대한 선제적인 조치, 최근 개정된 법령과 시행된 정책, 경찰교통민원 사이트, 단속 사후 절차에 대한 더욱 활발한 홍보가 필요한 것으로 판단된다.

• Journal of Information Processing Systems
• /
• 제14권1호
• /
• pp.140-149
• /
• 2018

Social data such as users' comments are unstructured in nature and up-to-date technologies for analyzing such data are constrained by the available storage space and processing time when fast storing and processing is required. On the other hand, it is even difficult in using a huge amount of dynamically generated social data to analyze the user features in a high speed. To solve this problem, we design and implement a topic association analysis system based on the latent Dirichlet allocation (LDA) model. The LDA does not require the training process and thus can analyze the social users' hourly interests on different topics in an easy way. The proposed system is constructed based on the Spark framework that is located on top of Hadoop cluster. It is advantageous of high-speed processing owing to that minimized access to hard disk is required and all the intermediately generated data are processed in the main memory. In the performance evaluation, it requires about 5 hours to analyze the topics for about 1 TB test social data (SNS comments). Moreover, through analyzing the association among topics, we can track the hourly change of social users' interests on different topics.

• 대한수학회보
• /
• 제53권6호
• /
• pp.1597-1612
• /
• 2016

There are fruitful results on degenerate parabolic-hyperbolic equations recently following the idea of $Kru{\check{z}}kov^{\prime}s$ doubling variables device. This paper is devoted to the well-posedness of nonhomogeneous boundary problem for degenerate parabolic-hyperbolic equations with spatially dependent second order operator, which has not caused much attention. The novelty is that we use the boundary flux triple instead of boundary layer to treat this problem.

• 한국경영과학회지
• /
• 제18권1호
• /
• pp.71-78
• /
• 1993

A Baysian approach is proposed is estimating the mission failure rates by criticalities. A mission failure which occurs according to a Poisson process with unknown rate is assumed to be classified as one of the criticality levels with an unknown probability. We employ the Gamma prior for the mission failure rate and the Dirichlet prior for the criticality probabilities. Posterior distributions of the mission rates by criticalities and predictive distributions of the time to failure are derived.

• ETRI Journal
• /
• 제43권1호
• /
• pp.74-81
• /
• 2021

The mixture model is a very powerful and flexible tool in clustering analysis. Based on the Dirichlet process and parsimonious Gaussian distribution, we propose a new nonparametric mixture framework for solving challenging clustering problems. Meanwhile, the inference of the model depends on the efficient online variational Bayesian approach, which enhances the information exchange between the whole and the part to a certain extent and applies to scalable datasets. The experiments on the scene database indicate that the novel clustering framework, when combined with a convolutional neural network for feature extraction, has meaningful advantages over other models.

• 정보과학회 논문지
• /
• 제44권10호
• /
• pp.1026-1033
• /
• 2017

본 논문은 계층적 디리슐레 과정(HDP)과 은닉 마르코프 모형(HMM)이 결합된 베이스 통계학적 방법과 HMM의 상태 지속 정보를 이용한 건강 상태 예측 방법을 제안한다. HDP-HMM은 베이스 방법의 HMM 확장 모형으로서 건강의 동적 특성을 고려하여 불확실하고 가늠하기조차도 어려운 건강 상태의 수를 추정할 수 있게 해준다. 모의 데이터와 실제 건건 검진 데이터를 이용한 시험을 통하여 흥미 있는 행동 특성을 볼 수 있었으며 최대 5년까지로 제한한 미래 예측도 충분한 가능함을 확인하였다. 미래는 불확실하며 예측 문제는 본질적으로 어렵다. 그러나 본 연구의 실험 결과로 동적인 문맥 하에서 다중 후보 가설을 제시함으로서 실용 가능한 건강상태의 장기 예측이 가능하다는 것을 읽을 수 있었다.

• 응용통계연구
• /
• 제16권1호
• /
• pp.101-115
• /
• 2003

본 논문에서는 Chen, Dey와 Shao(1999), Branco와 Dey(2001)가 제안한 왜도가 있는 두터운 꼬리를 가지는 오차 분포와 디리슈레 과정 사전분포를 이용한 베이지안 메타분석 (meta-analysis)을 하고자 한다. 베이지안 메타분석을 위하여 가중함수를 고려한 계층적 선택 모형을 이용한다. 이때의 오차항은 왜도가 있는 비정규 분포로 가정한다. 이를 위하여 우선 왜도 타원형 분포의 일반적인 족을 소개한다 이 분포족중 왜도 정규분포와 왜도 t 분포를 오차항 분포로 이용한 베이지안 계층적 선택 모형을 고려하며, 이 때 발생하는 복잡한 베이지안 계산은 MCMC 방법으로 해결한다. 마지막으로, 실제 자료(Johnson, 1993)인 두 가지의 충치예방약의 효과에 대한 차이를 비교하기 위해 얻어진 12개의 연구 자료를 이용하여 본 연구에서 제시된 베이지안 방법을 이용하여 메타분석을 한다.