• The KIPS Transactions:PartB
• /
• v.10B no.6
• /
• pp.587-592
• /
• 2003

Many real world control problems have continuous states and actions. When the state space is continuous, the reinforcement learning problems involve very large state space and suffer from memory and time for learning all individual state-action values. These problems need function approximators that reason action about new state from previously experienced states. We introduce Fuzzy Q-Map that is a function approximators for 1 - step Q-learning and is based on fuzzy clustering. Fuzzy Q-Map groups similar states and chooses an action and refers Q value according to membership degree. The centroid and Q value of winner cluster is updated using membership degree and TD(Temporal Difference) error. We applied Fuzzy Q-Map to the mountain car problem and acquired accelerated learning speed.

• Journal of the Korea Convergence Society
• /
• v.9 no.10
• /
• pp.415-421
• /
• 2018

The purpose of this study is to analyze the difference of scalp and hair care behaviors according to DiSC behaviors of beauty consumer. For the study, 251 questionnaires were analyzed using the SPSS v.18.0 statistical package program. First, there was statistically significant difference in psychological stability, favorable image, and maintaining state of scalp and hair management according to DiSC behavior type. In case of Shin Jung - hyung (C), psychological stability, Positive images, and maintaining good condition. Second, DiSC and hair management behaviors according to DiSC behaviors showed statistically significant difference in state cognition management and product use management. In the case of cognition type (C) and social type (I) Product use management. The results of this study suggest that hair management awareness and behavior, which is dependent on DiSC behaviors of consumers, are provided as basic data for corporate marketing strategy.

• Proceedings of the Korean Society for Emotion and Sensibility Conference
• /
• 2009.05a
• /
• pp.217-218
• /
• 2009

총 19 명 (male=9, female=10)을 대상으로 자기 보고식 질문지를 일주일 간격으로 2 회 수행한 후 double blind 방식과 실험 목적/프로토콜을 피검자에게 알려주지 않은 상태로 EEG 검사를 수행하여 행동활성화체계/행동억제체계(BAS/BIS)를 분석하였다. 안정기를 거친 후 갑자기 불유쾌한 청각적 자극(1Hz, 110dB, 15초간)을 2 회 실시하여 BAS/BIS 수치의 변화를 분석하였다. 분노활성화 음성자극에 대한 BAS/BIS 성향 대상자들의 반응을 분석하여 그 변화추이를 살펴보고자 함이 본 연구의 목적이다.

• Proceedings of The Korean Society of Health Promotion Conference
• /
• 2001.09a
• /
• pp.11-20
• /
• 2001

"건강이란 질병이나 허약함이 없는 상태를 말함이 아니라, 신체적 정신적 사회적인 복지(Wellbeing)를 뜻하는 것"이라고 WHO(1948)는 정의하였다. 건강은 환경과 생활풍습(Life Style), 유전 그리고 의학적 치료 등에 의하여 결정되며 최근 WHO는 건강제도가 국민건강에 지배적인 역할을 한다고 주장한다. 이런 관점에서 건강제도가 국민 건강에 미친 총괄적인 성취도를 비교해 보면 한국이 전 세계에서 35위, 일본이 1위, 미국은 15위를 차지하고 있다는 통계가 나왔다. 개인의 건강은 지역사회의 건강과 밀접한 관계를 가지고 있으며, 각 지역사회의 건강상태는 국가의 건강상태를 결정짓는다. 그러므로 건강증진 교육과 사업, 연구는 건강상태에 영향력을 미치는 생활풍습과 행동의 변화와 환경조성에 초점을 두고 활성화되어야 한다. 아울러 국민 건강증진 사업을 성공시키기 위해서는 건강정책을 수립하는데 모든 국민의 적극적인 참여가 절대적으로 필요하다.대적으로 필요하다.

• Journal of the Korean Society of Fisheries and Ocean Technology
• /
• v.31 no.1
• /
• pp.74-83
• /
• 1995

To estimate the parameters of a mathematical model of fishes' swimming behavior, the behavior in a experimental water tank was observed and analyzed using the video monitoring system. The tank was equipped with vertical circulation system, and measured $3,500L\;{\times}\;1,500B\;{\times}\;1,000H\;mm$ at flow channel and $1,200L\;{\times}\;900B\;{\times}\;500H\;mm$ at observational part. Rainbow trout, salmo gairdnerii were used as experimental fishes. Their swimming behavior in the tank was observed by the monitoring system, and the positions of every individual were checked at 0.5 second intervals by the image processing of recorded pictures for 5 minutes. The mean swimming speed calculated from the time series data of positions of every individual ranged from 2.5BL cm/sec to 2.9BL cm/sec at the stagnated flow. The mean swimming speed of 10 individuals in a school increased according to the flow speed. The mean swimming depth ranged from 17 cm to 38 cm even though it changed irregularly at the stagnated flow and gradually became stable according to the increase of flow speed. In the present study, the mean distance of individuals from wall of the tank varied from 17.6cm to 21.4cm. The mean distance between the nearest individual varied from 0.4BL cm to 0.7BL cm when 10 individuals in a school were observed. The mean dimension of fish schools became enlarged in all directions according to increase in the number of individuals, and as flow speed increased the horizontal dimension of fish schools expanded while their vertical dimension decreased.

• Therapeutic Science for Rehabilitation
• /
• v.12 no.4
• /
• pp.81-96
• /
• 2023

Objective : This study aimed to examine the association between the structural and functional aspects of social networks and health-promoting behaviors in community-dwelling older adults. Methods : Social networks based on structural and functional aspects and health-promoting behaviors by lifestyle were measured in 226 community-dwelling adults aged 65 years and over. To analyze the collected data, an independent t-test, one-way ANOVA, Pearson's correlation analysis, and hierarchical regression were performed. Results : The network size was the highest in the friend network, and the frequency of contact and social support was the highest in the non-living family network. Health-promoting behaviors were highest for activities of daily living and lowest for productive and social activities. All subfactors of social networks showed significant positive correlations with health-promoting behaviors. Hierarchical regression showed social support from neighbors as the variable with a significant effect on health-promoting behaviors; gender and depression were also influencing factors, and this model showed 37% explanatory power. Conclusion : To promote healthy behaviors of older adultsin the community, the development of health promotion programs and related policies considering social networks centering on social support from neighbors is required.

• Journal of Korean Academy of Nursing
• /
• v.35 no.1
• /
• pp.143-153
• /
• 2005

Purpose: This study was aimed at finding the effects of cycled lighting on body weight, physiological variables and the behavioral state of LBWI (low birth weight infants) in the NICU. Method: The subjects were 20 LBWI at 2 NICUs. They were assigned to an experimental or control group which consisted of 10 subjects in each. Cycled light was applied to the experimental group for 10 days. Result: It was certified that the application of cycled lighting resulted in increased body weight and O2 saturation, and decreased heart rate of the LBWI. However, there was no effect in decrease of respiration and stabilization of the behavioral state. Conclusion: The application of cycled lighting might be a nursing intervention which would in turn have positive effects on the growth of LBWI.

• Proceedings of the Korean Institute of Intelligent Systems Conference
• /
• 2006.11a
• /
• pp.101-104
• /
• 2006

유비쿼터스 시대가 다가오면서 앞으로 가정 및 회사 등 인간이 거주하며 생활하는 공간에서의 좀 더 편리하고 효율적인 다양한 정보를 인간에게 인지시켜 줄 수 있는 환경이 구축되어야한다. 이를 기반으로 유비쿼터스 주변장치들의 네트워크와 인간에게 많은 정보와 편리성이 좀 더 효율적으로 이루어져야 할 것이다. 이를 위해 본 논문에서는 센서모듈에서 추출되는 데이터를 신경망과 퍼지 알고리즘을 사용해 동작인식의 패턴을 분류하여 인간의 사고를 움직임 파악한다. 이러한 패턴의 분류를 통해 홈네트워크 시스템과의 센서모듈의 통신제어가 가능하게 된다 이를 바탕으로 패턴이 분류된 행동들의 명령으로 미리 지정된 간단한 손동작으로 여러 가전기기라든지 홈네트워크 시스템의 제어방식을 더욱 간단히 제어하며, 인간의 건강상태를 파악함으로써 인간행동과 상태에 따른 유비쿼터스 환경의 제어가 이루어 질 수 있는 시스템을 제안한다.

• Proceedings of the Optical Society of Korea Conference
• /
• 2000.02a
• /
• pp.304-305
• /
• 2000

자연계에 존재하는 잡음은 대부분의 경우에 있어서 그 부정적인 면이 부각되어 왔고, 따라서 이를 없애거나 줄이려는 노력들이 이루어져 왔다. 하지만 최근 들어 비선형계에서 나타나는 비선형 공명(stochastic resonance), 특히 생체계에서 이루어지는 정보의 전달에서의 잡음의 긍정적인 역할은 많은 관심을 끌고 있고 많은 연구가 이루어지고 있다[1,2]. 또한 잡음이 비선형 비평형 동역학계의 행동에 영향을 미쳐서 원하는 상태로 이끌어 가거나 원하는 상태를 유지하는데 결정적인 역할을 하는 것이 알려지고 있다[3]. (중략)

• Journal of KIISE:Software and Applications
• /
• v.30 no.11
• /
• pp.1054-1061
• /
• 2003

Q-Learning, a representative algorithm of reinforcement learning, experiences repeatedly until estimation values about all state-action pairs of state space converge and achieve optimal policies. When the state space is high dimensional or continuous, complex reinforcement learning tasks involve very large state space and suffer from storing all individual state values in a single table. We introduce Q-Map that is new function approximation method to get classified policies. As an agent learns on-line, Q-Map groups states of similar situations and adapts to new experiences repeatedly. State-action pairs necessary for fine control are treated in the form of rule. As a result of experiment in maze environment and mountain car problem, we can achieve classified knowledge and extract easily rules from Q-Map