• 제목/요약/키워드: reward time

검색결과 168건 처리시간 0.03초

온라인 리서치 플랫폼의 종합적인 인공지능 활용 분석 연구 (A Study on the Comprehensive Artificial Intelligence Utilization Analysis of Online Research Platform)

  • 임혜원;남원석
    • 한국융합학회논문지
    • /
    • 제11권10호
    • /
    • pp.187-195
    • /
    • 2020
  • 본 연구는 근미래 기술 융복합 환경에서 온라인 리서치 업무의 개선 방법 모색을 목적으로 하며, 향후 인공지능 기술 응용 및 활용을 위한 기초적인 학문적 토대를 제시한다. 구체적인 연구 방법 및 결과는 다음과 같다. 첫째, 이론적 고찰을 통해 리서치의 패러다임 변화, 온라인 리서치 플랫폼, 인공지능 기술들을 살펴보고 연구의 근거를 확보했다. 둘째, 사례를 수집하여 온라인 리서치 플랫폼의 리워드, 실행, 시각화 3가지 유형을 도출하였다. 셋째, 사례 분석을 통해 종합적인 인공지능 활용 동향과 방향성을 도출했다. 연구 결과 인공지능 기술이 가장 많이 사용되고 있는 단계는 분석, 시각화, 실행, 섭외, 설계, 매칭 순서였으며, 인공지능의 활용 방향성은 크게 업무 효율화, 인간의 사고 보조, 리서치 방식의 변화로 나눌 수 있다. 그러나 리서치 단계 중 계획 단계에는 인공지능 비롯하여 제공되는 기능이 온라인 리서치 플랫폼에서 전반적으로 부족했으며, 사용자 및 제품의 다양화에 따라 리서치 계획이 중요해지고 있는 시점에서 향후 인공지능을 활용한 리서치 계획 방안 개발 및 연구가 필요한 실정이다. 본 연구는 온라인 리서치 플랫폼 방법의 개선에 긍정적인 영향을 미칠 것으로 기대하며, 향후 인공지능 기반의 리서치 플랫폼과 시스템 연구 개발의 기틀을 마련하고자 한다.

Effects of multiple herb formula SEC-22 supplementation on dietary intake, picky eating behaviors, and growth indices in thin preschool children

  • Kim, Kijoon;Lee, Joonsuk;Paik, Hee Young;Yoon, Jihyun;Ryu, Bongha;Shim, Jae Eun
    • Nutrition Research and Practice
    • /
    • 제9권4호
    • /
    • pp.393-399
    • /
    • 2015
  • BACKGROUND/OBJECTIVES: Thin children may have insufficient intake of energy and nutrients, resulting in reduced immune function and growth. This study aimed to identify the effects of multiple herb formula SEC-22 supplementation on growth, dietary changes, and picky eating behaviors in thin children. SUBJECTS/METHODS: A double-blind, randomized clinical trial was conducted on 79 children aged 2-5 years with poor appetites, BMI percentile < 25, and without any illness. Subjects were given either SEC-22 (n = 35) or placebo (n = 44) for 2 months and followed for an additional 2 months. Three-day dietary records, questionnaires on picky eating behaviors, and anthropometric measures were collected. RESULTS: Energy, carbohydrate intake, and feeding difficulty improved in both groups during the intervention period. However, changes were maintained only in the SEC-22 group after 2 months of follow-up post-supplementation. 'Frequency of trying to feed' was improved in the SEC-22 group compared to the placebo group after the first month of follow-up (P < 0.05). Intakes of potassium and thiamine were improved in the SEC-22 group compared to the placebo group after the first month of intervention (P < 0.05). 'Frequency of food reward', eating amount, and intakes of carbohydrate, potassium, and vitamin C showed significant improvement compared to the placebo group after the second month of follow-up (P < 0.05). CONCLUSIONS: These results suggest that SEC-22 supplementation can improve parental feeding difficulty resulting from insufficient eating amount or picky eating as well as increase nutrient intake in thin children. Although these improvements were observable at least 2 months after supplementation, effects beyond this time frame need to be confirmed.

NXT 로봇을 이용한 SVG 기반 실시간 드로잉 (SVG Based Realtime Drawing Using NXT Robot)

  • 장호연;류승택;박진완
    • 한국콘텐츠학회:학술대회논문집
    • /
    • 한국콘텐츠학회 2009년도 춘계 종합학술대회 논문집
    • /
    • pp.146-151
    • /
    • 2009
  • 현대 설치 예술 분야에서 피지컬 컴퓨팅(physical computing)을 이용한 작품의 사례가 많아지고 있다. 하지만 인터랙션(interaction)을 위한 로봇 사용이 아닌 인터페이스(interface)가 장착된 드로잉(drawing)의 도구로 로봇을 사용하여 작업한 예는 쉽게 찾아 볼 수가 없다. 본 논문에서는 사용자와 통신할 수 있는 드로잉 작업용 시스템 설계와 개발 과정에 대해 언급하고자 한다. 작업 환경을 구성하는 로봇으로는 레고(Lego) 사(社)에서 나온 마인드스톰(Mindstorm) 지능형 로봇 NXT 시스템을 이용하였고, 현실에서의 실제 드로잉 환경과 이를 예측하여 운동을 시뮬레이트(simulate)하는 가상 환경으로 구분하였다. 실제 환경에서 드로잉을 하기 위하여 NXT 시스템을 제어할 수 있도록 하는 아이커맨드(Icommand) 라이브러리(library)를 이용하였고, 가상 환경을 표현하기 위하여 이미지 표현이 쉬운 프로세싱(processing) 라이브러리를 이용하였다. 라인(line) 드로잉을 위하여 벡터(vector) 방식 SVG(Scalable Vector Graphics) 파일을 기반으로 이미지 정보를 얻어 표현하였다. 이 시스템은 블루투스(bluetooth) 연동으로 실시간 통신이 가능하여 사용자의 요구에 따라 원하는 이미지를 만들어 낼 수 있다. 이러한 모습은 이미지의 결과에 그치는 것이 아니라 드로잉을 하는 과정에서 하나의 퍼포먼스(performance)로 작용할 수가 있다.

  • PDF

간호행정자의 리더십 유형에 관한 연구;거래적, 변혁적 리더십을 중심으로 (Transactional and Transformational Leadership Styles of The Nurse Administrators)

  • 김문실;박현태
    • 간호행정학회지
    • /
    • 제3권1호
    • /
    • pp.5-15
    • /
    • 1997
  • Today's rapidly changing healthcare environment and increasingly professional nursing practice indicate that identifying leadership characteristic of nursing leaders and executives is a vital importance in today' s time and also mandate innovative leadership for nursing service. Therefore, the purpose of this study is to examined the transformational, transactional leadership styles of the Nurse Administrators. The sample consisted of sixteen mid-level nurse administrators, fifty head nurses of 5 General Hospital. Data for this study was collected from Sep. 20 to 29 by Bass' MLQ Questionnaire. The data was analyzed by frequency, percentage, one-way ANOVA. Major findings are as follows : Appropriate one-way ANOVA tests revealed that the differences for transformational and transactional leadership styles of nurse executives, mid-level nurse administrators as perceived by their immediate subordinates were statistically significant(P<.05). The scores of transformational and contingent reward behaviors were declined of the mid-level nurse administrators. The transactional scores of nurse administrators were lower than transformational ones, which is a desirable findings. The result of this study, the mid-level nurses administrators were perceived as the highest transformational leader by their subordinates. The nurse executives received the lower transformational leadership scores than mid-leval administrators. These results were opposit to the previous studies. Leader can aspire to these qualities of transformational leadership, building on the more traditional transactional dimensions. We can think that transformational leadership suggests a direction for developing a creative and rewarding approach to the leadership of professionalnursing practice environments. More research on transformational qualities in nursing service and controlled designs would be desirable for nursing service administration.

  • PDF

강화학습을 이용한 n-Queen 문제의 수렴속도 향상 (The Improvement of Convergence Rate in n-Queen Problem Using Reinforcement learning)

  • 임수연;손기준;박성배;이상조
    • 한국지능시스템학회논문지
    • /
    • 제15권1호
    • /
    • pp.1-5
    • /
    • 2005
  • 강화학습(Reinforcement-Learning)의 목적은 환경으로부터 주어지는 보상(reward)을 최대화하는 것이며, 강화학습 에이전트는 외부에 존재하는 환경과 시행착오를 통하여 상호작용하면서 학습한다 대표적인 강화학습 알고리즘인 Q-Learning은 시간 변화에 따른 적합도의 차이를 학습에 이용하는 TD-Learning의 한 종류로서 상태공간의 모든 상태-행동 쌍에 대한 평가 값을 반복 경험하여 최적의 전략을 얻는 방법이다. 본 논문에서는 강화학습을 적용하기 위한 예를 n-Queen 문제로 정하고, 문제풀이 알고리즘으로 Q-Learning을 사용하였다. n-Queen 문제를 해결하는 기존의 방법들과 제안한 방법을 비교 실험한 격과, 강화학습을 이용한 방법이 목표에 도달하기 위한 상태전이의 수를 줄여줌으로써 최적 해에 수련하는 속도가 더욱 빠름을 알 수 있었다.

Altered Functional Disconnectivity in Internet Addicts with Resting-State Functional Magnetic Resonance Imaging

  • Seok, Ji-Woo;Sohn, Jin-Hun
    • 대한인간공학회지
    • /
    • 제33권5호
    • /
    • pp.377-386
    • /
    • 2014
  • Objective: In this study, we used resting-state fMRI data to map differences in functional connectivity between a comprehensive set of 8 distinct cortical and subcortical brain regions in healthy controls and Internet addicts. We also investigated the relationship between resting state connectivity strength and the level of psychopathology (ex. score of internet addiction scale and score of Barratt impulsiveness scale). Background: There is a lot of evidence of relationship between Internet addiction and impaired inhibitory control. Clinical evidence suggests that Internet addicts have a high level of impulsivity as measured by behavioral task of response inhibition and a self report questionnaire. Method: 15 Internet addicts and 15 demographically similar non-addicts participated in the current resting-state fMRI experiment. For the connectivity analysis, regions of interests (ROIs) were defined based on the previous studies of addictions. Functional connectivity assessment for each subject was obtained by correlating time-series across the ROIs, resulting in $8{\times}8$ matrixs for each subject. Within-group, functional connectivity patterns were observed by entering the z maps of the ROIs of each subject into second-level one sample t test. Two sample t test was also performed to examine between group differences. Results: Between group, the analysis revealed that the connectivity in between the orbito frontal cortex and inferior parietal cortex, between orbito frontal cortex and putamen, between the orbito frontal cortex and anterior cingulate cortex, between the insula and anterior cingulate cortex, and between amydgala and insula was significantly stronger in control group than in the Internet addicts, while the connectivity in between the orbito frontal cortex and insula showed stronger negative correlation in the Internet addicts relative to control group (p < 0.001, uncorrected). No significant relationship between functional connectivity strength and current degree of Internet addiction and degree of impulsitivy was seen. Conclusion: This study found that Internet addicts had declined connectivity strength in the orbitofrontal cortex (OFC) and other regions (e.g., ACC, IPC, and insula) during resting-state. It may reflect deficits in the OFC function to process information from different area in the corticostriatal reward network. Application: The results might help to develop theoretical modeling of Internet addiction for Internet addiction discrimination.

초혼계모의 재혼가족생활 스트레스와 적응에 대한 경험적 연구 (The Experience of the first-married stepmother's stress and adaptation to her stepfamily)

  • 임춘희
    • 대한가정학회지
    • /
    • 제35권5호
    • /
    • pp.73-102
    • /
    • 1997
  • The purpose of this study was to identify the specific and empirical contents in terms of first-married stepmother about her stress and efforts to adapt to her stepfamily. This studies relied on the qualitative research method in order to approach stepmother's subjective experience. There were 6 stepmother respondents who were 6 first-marrieds living in Seoul or Kyoungki area. The data was collected by in-depth personal interview.‘Open coding process’, a part of grounded theory procedures were used to analyzed the data. The major findings can be summarized as follows: 1. When focused o stressors, the stress of stepmothers were categorized into 4 categories. Those were the following: 1 When focused on stressors, the stress of stepmothers were categorized into 4 categories. Those were the following: (1) Stepmother's own problem such as stepmother' identity problem; (2) Problems within stepfamily such as problems about relations with husbands, stepchildren, and her own children; (3) Problems of stepmother family's kin such as relational problems with former spouses, in-laws, and her parent; (4) Problems of social relationship such as isolation and alienation in social relationship and neighbors' prejudice about stepmother. 2. In overcoming stress from stepfamily life and adapting to it, 4 categories were identified as important. Contents of 4 categories were such like this: (1) Stepmother's own efforts - (a) give a positive meaning to marriage or remarriage (b) flexible personality (c) situation-accepting attitude (d) exclusive management of household income (e) job (f) perseverance for time to adaptation; (2) Positive feedback within stepfamily - (a) supportive husband (b) sense of unity as a couple (c) psychological reward from stepchildren (d) existence of stepmother's own child (e) behavior for family solidarity (f) family-centered religious life; (3) Stepfamily's relationship with former spouses such as breakdown in relationship with ex-wives or ex-husbands; (4) Support from stepmothers' kin and people in social network with stepmothers. It was concluded that to reduce stress and to help adapt to stepfamily, stepmother should make an effort, and that her stepfamily and her kin should cooperate and support. In addition, negative stereotype of stepmother and stepfamily should disappeared. It was suggested that studies on stepfamily should be vitalized and that programs for preparing remarriage and counseling for stepfamily should be spread.

  • PDF

상관차원에 의한 비선형 뇌파 분석과 기질성격척도(TCI) 요인간의 상관분석 (Correlation over Nonlinear Analysis of EEG and TCI Factor)

  • 박진성;박영배;박영재;허영
    • 대한한의진단학회지
    • /
    • 제11권2호
    • /
    • pp.96-115
    • /
    • 2007
  • Background and Purpose: Electroencephalogram(EEG) is a multi-scaled signal consisting of several components of time series with different origins. Recently, because of the absence of an identified metric which quantifies the complex amount of information, there are many limitations in using such a linear method. According to chaos theory, irregular signals of EEG can also result from low dimensional deterministic chaos. Chaotic nonlinear dynamics in the EEG can be studied by calculating the correlation dimension. The aim of this study is to analyze correlation between the correlation dimension of EEG and psychological Test (TCI). Methods: Before and after moxibustion treatment, EEG raw data were measured by moving windows during 15 minutes. The correlation dimension(D2) was calculated from stabilized 40 seconds in 15 minutes data. 8 channels EEG study on the Fp, F, T, P was carried out in 30 subjects. Results: Correlation analysis of TCI test is calculated with deterministic non-linear data and stochastic non-linear data. 1. Novelty seeking in temperament is positive correlated with D2 of EEG on Fp. 2. reward dependence in temperament is positive correlated with D2 of EEG on T3,T4 and negative correlated with D2 of EEG on P3,P4. 3. self directedness in character is positive correlated with D2 of EEG on F4, P3. 4. Harm avoidance is negative correlated with D2 of EEG on Fp2, T3, P3. Conclusion: These results suggest that nonlinear analysis of EEG can quantify dynamic state of brain abolut psychological Test (TCI).

  • PDF

포스트 코로나19 시대의 패션 소비문화에 대한 빅데이터 분석 -중국 패션 네트워크인 LADYMAX.cn의 기사를 중심으로- (Fashion Consumption Culture in the Post-COVID-19 Era Identified through Big Data Analysis -Focusing on Articles in the Chinese Fashion Network LADYMAX.cn-)

  • 빈삼;염혜정;심수인
    • 패션비즈니스
    • /
    • 제25권2호
    • /
    • pp.80-97
    • /
    • 2021
  • In this study, the changes in fashion consumption culture in the post-COVID-19 era were examined through big data analysis. Considering that the Chinese market plays a pivotal role in the global fashion industry, big data was collected in the most famous and professional fashion network in China, LADYMAX.cn. As a result of text mining and social network analysis, three major changes were identified as the emerging fashion consumption culture in the post-COVID-19 era. First, as a trend in new media consumption, COVID-19 disease and the development of digital technology tended to encourage consumers to put more importance on the relationship between bloggers and fans than previously. Second, as a trend in reward consumption, consumers tended to be rewarded for their hard work to relieve and comfort their high stress caused by spending a long time worrying about the prolonged COVID-19 situation. Third, as a trend in home-economy consumption, consumers tended to prefer homewear and sportswear more because they were spending longer times at home as the social distancing period was prolonged.

Deep Q 학습 기반의 다중경로 시스템 경로 선택 알고리즘 (Path selection algorithm for multi-path system based on deep Q learning)

  • 정병창;박혜숙
    • 한국정보통신학회논문지
    • /
    • 제25권1호
    • /
    • pp.50-55
    • /
    • 2021
  • 다중경로 시스템은 유선망, LTE망, 위성망 등 다양한 망을 동시에 활용하여 데이터를 전송하는 시스템으로, 통신망의 전송속도, 신뢰도, 보안성 등을 높이기 위해 제안되었다. 본 논문에서는 이 시스템에서 각 망의 지연시간을 보상으로 하는 강화학습 기반 경로 선택 방안을 제안하고자 한다. 기존의 강화학습 모델과는 다르게, deep Q 학습을 이용하여 망의 변화하는 환경에 즉각적으로 대응하도록 알고리즘을 설계하였다. 네트워크 환경에서는 보상 정보를 일정 지연시간이 지나야 얻을 수 있으므로 이를 보정하는 방안 또한 함께 제안하였다. 성능을 평가하기 위해, 분산 데이터베이스와 텐서플로우 모듈 등을 포함한 테스트베드 학습 서버를 개발하였다. 시뮬레이션 결과, 제안 알고리즘이 RTT 감소 측면에서 최저 지연시간을 선택하는 방안보다 20% 가량 좋은 성능을 가지는 것을 확인하였다.