• Communications for Statistical Applications and Methods
• /
• 제23권2호
• /
• pp.131-146
• /
• 2016

In research on behavioral studies, significant attention has been paid to the stage-sequential process for longitudinal data. Latent class profile analysis (LCPA) is an useful method to study sequential patterns of the behavioral development by the two-step identification process: identifying a small number of latent classes at each measurement occasion and two or more homogeneous subgroups in which individuals exhibit a similar sequence of latent class membership over time. Maximum likelihood (ML) estimates for LCPA are easily obtained by expectation-maximization (EM) algorithm, and Bayesian inference can be implemented via Markov chain Monte Carlo (MCMC). However, unusual properties in the likelihood of LCPA can cause difficulties in ML and Bayesian inference as well as estimation in small samples. This article describes and addresses erratic problems that involve conventional ML and Bayesian estimates for LCPA with small samples. We argue that these problems can be alleviated with a small amount of prior input. This study evaluates the performance of likelihood and MCMC-based estimates with the proposed prior in drawing inference over repeated sampling. Our simulation shows that estimates from the proposed methods perform better than those from the conventional ML and Bayesian method.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제5권11호
• /
• pp.2016-2034
• /
• 2011

Ubiquitous Healthcare (u-Healthcare) is the intelligent delivery of healthcare services to users anytime and anywhere. To provide robust healthcare services, recognition of patient daily life activities is required. Context information in combination with user real-time daily life activities can help in the provision of more personalized services, service suggestions, and changes in system behavior based on user profile for better healthcare services. In this paper, we focus on the intelligent manipulation of activities using the Context-aware Activity Manipulation Engine (CAME) core of the Human Activity Recognition Engine (HARE). The activities are recognized using video-based, wearable sensor-based, and location-based activity recognition engines. An ontology-based activity fusion with subject profile information for personalized system response is achieved. CAME receives real-time low level activities and infers higher level activities, situation analysis, personalized service suggestions, and makes appropriate decisions. A two-phase filtering technique is applied for intelligent processing of information (represented in ontology) and making appropriate decisions based on rules (incorporating expert knowledge). The experimental results for intelligent processing of activity information showed relatively better accuracy. Moreover, CAME is extended with activity filters and T-Box inference that resulted in better accuracy and response time in comparison to initial results of CAME.

• Asian-Australasian Journal of Animal Sciences
• /
• 제13권8호
• /
• pp.1035-1039
• /
• 2000

This study developed a Poisson generalized linear mixed model and a procedure to estimate genetic parameters for count traits. The method derived from a frequentist perspective was based on hierarchical likelihood, and the maximum adjusted profile hierarchical likelihood was employed to estimate dispersion parameters of genetic random effects. Current approach is a generalization of Henderson's method to non-normal data, and was applied to simulated data. Underestimation was observed in the genetic variance component estimates for the data simulated with large heritability by using the Poisson generalized linear mixed model and the corresponding maximum adjusted profile hierarchical likelihood. However, the current method fitted the data generated with small heritability better than those generated with large heritability.

• 한국정보과학회논문지:소프트웨어및응용
• /
• 제33권12호
• /
• pp.1062-1072
• /
• 2006

최근, '유비쿼터스 컴퓨팅'이라는 지능형 서비스 프레임워크가 제안되면서 적응형 에이전트 시스템의 필요성이 점점 증가되기 시작됐다. 본 논문에서는 유비쿼터스 컴퓨팅 시스템이 사용자에게 적절한 서비스를 제공하도록 도와주는 지능형 서비스 에이전트를 제안한다. 사용자에게 적절한 유비쿼터스 서비스를 제공하기 위해서는, 각각의 유비쿼터스 서비스 시스템 내에서의 상황 정보(Context Information) 차이를 조절하고 사용자의 취향을 서비스에 반영해야 한다. 따라서 다음 3가지 부분에 중점을 두어 연구를 진행하였다. 첫째, 적절한 다중 에이전트 프레임워크-에이전트간의 커뮤니케이션 이해와 추론엔진의 적용, 둘째, 유비쿼터스 컴퓨팅 환경 내에 존재하는 다양한 상황 정보(Context information)를 효과적으로 표현하는 유비쿼터스 온톨로지-에이전트간의 상황 정보 공유와 이해, 마지막으로 유비쿼터스 시스템에 적용되는 사용자 프로파일 구축 방법에 대해 연구 하였다. 본 논문에서 제안하는 지능형 서비스 에이전트는 사용자 취향에 따라 적절한 서비스를 제공하는 적응형 유비쿼터스 서비스 시스템 구축을 가능하게 한다.

• 한국정보통신학회:학술대회논문집
• /
• 한국정보통신학회 2014년도 춘계학술대회
• /
• pp.901-904
• /
• 2014

센싱 기술이 발달함에 따라, 다양한 도메인에서 각각 다른 목적을 가진 상황인지 기술들이 개발되어 왔다. 많은 서비스들이 개발되고 실제로 사용되고 있지만, 특정한 공간을 도메인으로 둔 서비스는 미흡하였다. 이를 해결하고자 여러 시도들이 있어왔으며 실제로 성과를 거둔 연구도 많이 있었다. 이 중, 온톨로지 기반 추론을 이용한 방법들이 상황인지 시스템에 비교적 신뢰성 있고 적합하였으나, 복잡한 규칙을 사용하지 않는 이상 사용자 별로 상황을 인지하지는 못하였다. 본 논문에서는 기존의 사용자 중심의 상황 모델링에 기반 한 추론에서 벗어나 개체 중심으로 그 범위를 확대하였다. 또한, 사용자 및 홈 프로파일을 사용하여 사용자 별 상황을 인지 할 뿐만 아니라 각 개체 별 상황을 인지함으로써 더욱 구체적인 상황정보를 제공하고자 한다.

• 한국지능시스템학회:학술대회논문집
• /
• 한국퍼지및지능시스템학회 1993년도 Fifth International Fuzzy Systems Association World Congress 93
• /
• pp.1370-1373
• /
• 1993

In controlling a system having many variables to control and multi objectives to satisfy such as a roof crane system, it is often difficult to obtain fuzzy If-Then rules in usual ways. As an alternative, we can more easely obtain rules in such a manner that we obtain each independent group of rules using partial variables for a partial objective. In this case, obtained rules can be conflicting with each other and conventional inference methods cannot handle such rules effectively. In this paper, we propose a roof crane controller with optimal velocity profile generator and a fuzzy logic controller with an inference method suitable for such conflicting rules.

• 한국정보과학회논문지:시스템및이론
• /
• 제33권10호
• /
• pp.709-721
• /
• 2006

지상파, 위성파, 케이블 방송 같은 기존의 방송환경은 시청자 취향에 상관없이 일방적인 단방향 방송 서비스를 제공해 왔다. 하지만, 최근에는 광대역 통신망을 통한 다양한 미디어 전송이 가능하게 되었다. 또한, 방송 환경에서 양방향 통신이 가능하게 됨으로써 장르, 시청 시간대, 배우 등 시청자의 선호도를 반영한 방송 서비스가 중용한 응용으로 대두되고 있다. 따라서, 기존의 방송환경에서 시청자의 선호도를 반영한 맞춤형 방송 서비스가 중요한 방송 서비스의 하나가 될 수 있다. 본 논문에서는 표적광고를 위한 새로운 시도로써 맞춤형 방송 서비스 응용 중 하나인 시청자 프로과일 추론을 통한 표적 광고 방법을 소개한다. 제안된 시청자 프로파일 추론 알고리즘은 시청자의 TV 시청 데이타(TV Viewing history data) 분석을 통해 시청자의 성별 및 연령대를 예측한다. 예측된 시청자의 성별 및 연령대를 바탕으로 TV Anytime 메타데이타를 이용한 표적 광고 선별 방법을 통하여 광고를 선택하게 된다. 제안된 표적 광고 시스템은 시청자 프로파일 추론 알고리즘과 표적 광고 선별 방법을 이용하여 구성되어 있으며, 실제 TV 시청 데이타를 이용하여 제안된 표적 광고 시스템의 실험 결과를 제시한다.

• Journal of the Korean Data and Information Science Society
• /
• 제15권1호
• /
• pp.251-261
• /
• 2004

The paper focuses primarily on the standard linear multiple regression model where the parameter of interest is a ratio of two regression coefficients. The general model includes the calibration model, the Fieller-Creasy problem, slope-ratio assays, parallel-line assays, and bioequivalence. We provide an orthogonal transformation (cf. Cox and Reid (1987)) of the original parameter vector. Also, we give some remarks on the difficulties associated with likelihood based confidence interval.

• Communications for Statistical Applications and Methods
• /
• 제27권2호
• /
• pp.241-254
• /
• 2020

This study proposes generalized models of joint latent class analysis (JLCA) for longitudinal data in two approaches, a JLCA with latent profile (JLCPA) and a JLCA with latent transition (JLTA). Our models reflect cross-sectional as well as longitudinal dependence among multiple latent classes and track multiple class-sequences over time. For the identifiability and meaningful inference, EM algorithm produces maximum-likelihood estimates under local independence assumptions. As an empirical analysis, we apply our models to track the joint patterns of adolescent depression and anxiety among US adolescents and show that both JLCPA and JLTA identify three adolescent emotional well-being subgroups. In addition, JLCPA classifies two representative profiles for these emotional well-being subgroups across time, and these profiles have different tendencies according to the parent-adolescent-relationship subgroups.

• 한국환경과학회지
• /
• 제13권3호
• /
• pp.245-250
• /
• 2004

The SBR(Sequencing Batch Reactor) process is ideally suited to treat high loading wastewater due to its high dilution rate. SBR operates by a cycle of periods consisting of filling, reacting, settling, decanting and idling. The react phases such as aeration or non-aeration, organic oxidation, nitrification, denitrification and other bio-logical reactions can be achieved in a reactor. Although the whole reactions can be achieved in a SBR with time distributing, it is hard to manage the SBR as a normal condition without recognizing a present state. The present state can be observed with nutrient sensors such as ${NH_{4}}^{+}-N$, ${NO_{2}}^{-}-N$, ${NO_{3}}^{-}-N} and ${PO_{4}}^{ 3-}-P.$ However, there is still a disadvantage to use the nutrient sensors because of their high expense and inconvenience to manage. Therefore, it is very useful to use common on-line sensors such as DO, ORP and pH, which are less expensive and more convient. Moreover, the present states and unexpected changes of SBR might be predicted by using of them. This study was conducted to get basic materials for making an inference of SBR process from ORP(oxidation reduction potential) of synthetic wastewater. The profiles of ORP, DO, and pH were under normal nitrification and denitrification were obtained to compare abnormal condition. And also, nitrite and nitrate accumulation were investigated during reaction of SBR. The bending point on ORP profile was not entirely in the low COD/NOx ratio condition. In this case, NOx was not entirely removed, and minimum ORP value was presented over -300mV. Under suitable COD/NOx ratio which complete denitrification was achieved, ORP bending point was observed and minimum ORP value was under -300m V. Under high COD/NOx ratio, ORP bending point was not detected at the first subcycle because of the fast denitrification and minimum ORP value was under -300mV at the time.