• Proceedings of the Korean Information Science Society Conference
• /
• 2004.10a
• /
• pp.94-96
• /
• 2004

에이전트의 '물체 따라가기'는 전통적으로 자동운전이나 가이드 등의 다양한 서비스를 제공할 수 있는 기본적인 기능이다. 여러 가지 물체가 있는 환경에서 '물체 따라가기'를 하기 위해서는 목적하는 대상이 어디에 있는지 찾을 수 있어야 하며, 실제 환경에는 사람이나 차와 같이 움직이는 물체들이 존재하기 때문에 다른 물체들을 피할 수 있어야 한다. 그런데 에이전트의 최적화된 피하기 행동은 장애물의 모양과 크기에 따라 다르게 생성될 수 있다. 본 논문에서는 다양한 모양과 크기의 장애물이 있는 환경에서 최적의 피하기 행동을 생성하면서 물체를 추적하기 위해 반응형 에이전트의 행동선택을 강화학습 한다. 여기에서 정확하게 상태를 인식하기 위하여 상태를 추론하고 목표물과 일정거리를 유지하기 위해 베이지안 추론을 이용한다 베이지안 추론은 센서정보를 이용해 확률 테이블을 생성하고 가장 유력한 상황을 추론하는데 적합한 방법이고, 강화학습은 실시간으로 장애물 종류에 따른 상태에서 최적화된 행동을 생성하도록 평가함수를 제공하기 때문에 베이지안 추론과 강화학습의 결합모델로 장애물에 따른 최적의 피하기 행동을 생성할 수 있다. Webot을 이용한 시뮬레이션을 통하여 다양한 물체가 존재하는 환경에서 목적하는 대상을 따라가면서 이종의 움직이는 장애물을 최적화된 방법으로 피할 수 있음을 확인하였다.

• Proceedings of the Korean Information Science Society Conference
• /
• 2001.10b
• /
• pp.55-57
• /
• 2001

강화학습은 한 에이전트가 자신이 놓여진 환경으로부터의 보상을 최대화할 수 있는 최적의 행동 전략을 학습하는 것이다. 따라서 강화학습은 입력(상태)과 출력(행동)의 쌍으로 명확한 훈련 예들이 제공되는 교사 학습과는 다르다. 특히 Q-학습과 같은 비 모델 기반(model-free)의 강화학습은 사전에 환경에 대한 별다른 모델을 설정하거나 학습할 필요가 없으며 다양한 상태와 행동들을 충분히 자주 경험할 수만 있으면 최적의 행동전략에 도달할 수 있어 다양한 응용분야에 적용되고 있다. 하지만 실제 응용분야에서 Q-학습과 같은 강화학습이 겪는 최대의 문제는 큰 상태 공간을 갖는 문제의 경우에는 적절한 시간 내에 각 상태와 행동들에 대한 최적의 Q값에 수렴할 수 없어 효과를 거두기 어렵다는 점이다. 이런 문제점을 고려하여 본 논문에서는 로봇 축구 시뮬레이션 환경에서 각 선수 에이전트의 동적 위치 결정을 위해 효과적인 새로운 Q-학습 방법을 제안한다. 이 방법은 원래 문제의 상태공간을 몇 개의 작은 모듈들로 나누고 이들의 개별적인 Q-학습 결과를 단순히 결합하는 종래의 모듈화 Q-학습(Modular Q-Learning)을 개선하여, 보상에 끼친 각 모듈의 기여도에 따라 모듈들의 학습결과를 적응적으로 결합하는 방법이다. 이와 같은 적응적 중재에 기초한 모듈화 Q-학습법(Adaptive Mediation based Modular Q-Learning, AMMQL)은 종래의 모듈화 Q-학습법의 장점과 마찬가지로 큰 상태공간의 문제를 해결할 수 있을 뿐 아니라 보다 동적인 환경변화에 유연하게 적응하여 새로운 행동 전략을 학습할 수 있다는 장점을 추가로 가질 수 있다. 이러한 특성을 지닌 AMMQL 학습법은 로봇축구와 같이 끊임없이 실시간적으로 변화가 일어나는 다중 에이전트 환경에서 특히 높은 효과를 볼 수 있다. 본 논문에서는 AMMQL 학습방법의 개념을 소개하고, 로봇축구 에이전트의 동적 위치 결정을 위한 학습에 어떻게 이 학습방법을 적용할 수 있는지 세부 설계를 제시한다.

• Journal of Navigation and Port Research
• /
• v.43 no.4
• /
• pp.237-242
• /
• 2019

Behavioral errors of the seafarers are one of the major causes of collisions and are usually corrected through education and training. To correct this behavioral error, the structure in which the behavioral error occurs needs to be identified and analyzed. For this purpose, behavior observation data were obtained through ship maneuvering simulation for collision encounters. The 9-state behavior classification frame proposed by Reason was used for the behavior observation and 50 university students were involved in the experiment. Behavioral analysis used the behavioral model of collision avoidance success and failure, which was developed from the 9-state Left-to-Right Hidden Markov modeling technique. As a result of the experiment, the difference between behaviors of success and failure of collision avoidance was clearly identified, and the linkage between 9-state behaviors, required to prevent collision, was derived.

• 한국노년학
• /
• v.32 no.2
• /
• pp.593-609
• /
• 2012

The study was performed to identify the functional status change of beneficiaries of Long-Term Care Insurance and its related factors. We conducted the logistic regression with 17,652 beneficiaries during August and September in 2008. As a result, activities of daily living(ADL), behavioral changes, rehab, instrumental activities of daily living(IADL) and cognitive function, followed by nursing care area were improved in a greater degree. For the institutional service, level-1 beneficiaries was significantly improved in rehab area and level-2 beneficiaries was improved in ADL. For the home-visit care service of in-home services, level-1 beneficiaries was improved in ADL, level-2 beneficiaries was improved in ADL and rehab area, level-3 beneficiaries was improved in ADL, cognitive function and behavioral changes. For the day-and-night care service, level-1 beneficiaries was improved in ADL, IADL, behavioral changes and rehab area, level-2 beneficiaries was improved in behavioral changes, level-3 beneficiaries was improved in cognitive function and behavioral changes. For the short-stay service, level-3 beneficiaries was improved in behavioral changes. By the above results, there was a difference in a functional improvement by level and used services. Therefore, government need to provide the personalized service system based on the objective and comprehensive understanding for health and functional status of beneficiaries.

• Journal of Digital Convergence
• /
• v.13 no.6
• /
• pp.287-296
• /
• 2015

This study was purposed to figure out current situation of part of 273 middle-school girls' dietary behavior and problems, then provide baseline data which helps to develop desirable eating behavior and life style. It appeared that suitable amount of eating(F=6.378, p=.002), degree of unbalanced diet(F=3.542, p=.030), Nutrients intake(F=6.780, p=.001), and life style(F=4.291, p=.015) differs between groups depending on student's subjective physical health status. It turned up that perceived physical health status is related to moderate amount of eating, degree of unbalanced diet, and Nutrients intake(p<0.05). As a results, in the period of adolescent which is a rapid growth period, there should be a systematic and endurable practice of right eating behavior, foods for formation of lifestyle, and convergence education in health area.

• Journal of Internet Computing and Services
• /
• v.12 no.2
• /
• pp.9-17
• /
• 2011

This paper proposes an algorithm that predicts the state of user's next actions, exploiting the HMM (Hidden Markov Model) on user profile data stored in the ubiquitous home network. The HMM, recognizes patterns of sequential data, adequately represents the temporal property implicated in the data, and is a typical model that can infer information from the sequential data. The proposed algorithm uses the number of the user's action performed, the location and duration of the actions saved by "Activity Recognition System" as training data. An objective formulation for the user's interest in his action is proposed by giving weight on his action, and change on the state of his next action is predicted by obtaining the change on the weight according to the flow of time using the HMM. The proposed algorithm, helps constructing realistic ubiquitous home networks.

• Journal of Digital Convergence
• /
• v.18 no.1
• /
• pp.321-332
• /
• 2020

The purpose of this study was to identify relationships among level of e-Health literacy, subjective health status and health-seeking behaviors. Subjects were 102 senior citizens who used the internet at least once in the last month and visited elderly community centers and welfare centers. Collected data were analyzed descriptive statistics, t-test, ANOVA, Pearson's Correlation Coefficient, and Multiple Regression using SPSS version 25.0. Fifty five percent of subjects responded the average weekly internet usage hours was 'less than 3 hours', The e-health literacy (r=.349, p<.001) and subjective health status (r=.466, p<.001) were positively correlated with health seeking behavior. Subjects' gender (β=.22, p=.012), type of disease (β=.23, p=.013) and subjective health status (β=.38, p<.001) had significant effects on health-seeking behaviors. The results suggest that the development of intervention programs to increase the level of e-Health literacy and subjective health status to improve of their health-seeking behaviors.

• Journal of Convergence for Information Technology
• /
• v.9 no.4
• /
• pp.118-129
• /
• 2019

This study aimed to analyze the association among stress, sugar intake behaviors and subjective oral health status and provide basic data for oral health care program related to nutrition education to practice reducing sugar intake. The subjects of this study were 273 high school students, research tool was structured, anonymous, self-administrated questionnaire. The data were analyzed using SPSSWIN version 18.0. Significant positive correlation was found between stress and sugar intake behaviors. Significant negative correlation was between stress and subjective oral health status and between sugar intake behaviors and subjective oral health status. Stress and sugar intake behaviors are related to oral health status. So it is very important to develop oral health program in order to improve eating habit and oral health status of high school students.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2005.05a
• /
• pp.325-328
• /
• 2005

기존의 다사용자 온라인 게임에서 구현되어 있는 MOB(Mobile Character)들은 '대기' 와 '공격' 의 2 가지 상태를 가지며 사용자의 '공격' 이라는 이벤트에만 반응하도록 설계되어 있는 수동적인 에이전트들이다. 본 논문에서는 기존의 '대기' 와 '공격' 상태에 '회피' 상태를 추가하고 3 가지 각각의 행동 전이에 따른 행동 패턴을 행동 특성 곡선으로 표현하며 '공격' 과 '접근' 자극을 스트레스 모형에 적용하여 스트레스에 따른 MOB 에이전트의 행동 패턴 변화를 설명하고 주변의 다른 에이전트들과의 협동을 도모할 수 있는 지능적인 NPC 에이전트를 설계한다.

• Korean Journal of Psychosomatic Medicine
• /
• v.11 no.2
• /
• pp.205-213
• /
• 2003

Objectives: The authors intended to evaluate long-term outcome of group cognitive behavioral therapy(GCBT) for panic disorder and examined the variables to predict high end-state functioning in 12 months follow-up. Methods: 236 patients meeting DSM-IV criteria for panic disorder were assessed by STAI, ASI, BDI, BSQ, PBQ, ACQ at pre & post treatment, and were asked about the frequency of panic attacks during recent one month. We executed stepwise discriminant analysis on the clinical variables at pre treatment assessment to find the variables for discriminating between high end-state function(HES) and low end-state function(LES). Results: After GCBT, 82.6% of panic patients maintained high end-state functioning at 12 months follow-up. The results of discriminant analysis showed that BDI and BSQ at pre treatment assessment were significant variables to predict end-state functioning at 12 months follow-up. The hit ratio of discriminant analysis was 69.2%. Conclusion: These results suggest that the therapeutic effect of GCBT can be maintained through 12 months. Especially, it is likely that patients who were less depressed and who had less experienced the fear of physical symptoms at pre-treatment can maintain HES.