Earticle

본 논문은 스마트폰을 이용하여 장시간동안 연속적으로 사용자의 행동정보를 수집 및 기록하는데 있어 요구되는 최적화된 수집주기를 제안한다. 스마트폰에서 장시간 기록을 하려면 배터리소모, 데이터 크기의 증가, 그리고 고성능 컴퓨팅의 요구와 같은 문제점을 고려해야 하며, 이러한 문제점은 수집주기를 최적화함으로써 해결가능하다. 본 논문에서는, 적절한 정확도를 유지하는 최적화된 수집주기를 제안하기 위해 실제 일상생활에서 스마트폰을 이용하여 사용자의 행동을 수집하고 기록하는 실험을 진행하였다. 실험결과 앉기의 경우 32초, 걷기와 서기의 경우 8초의 수집주기 내에서 약 80% 이상의 분류 정확성을 보이고 있음을 확인하였다. 결론적으로 앉기, 걷기, 서기와 같은 정도의 사용자의 행동을 일상생활에서 기록하기 위한 최적의 수집주기는 8초로 고려된다. 이러한 결론은 스마트폰을 이용한 디지털 문진표에 사용될 수 있다.

In this paper, we propose optimized data collection cycle required for long-term and continues collecting and recording of user’s activity information in smartphone environment. For long-term recording in smartphones, we face 3 problems such as battery durability, data length, and computing power. These problems can be solved by optimizing data collection cycle. To propose optimal data collection cycle while maintaining required accuracy, we conduct experiments by collecting and recording daily activity of a user through a smartphone. The experimental results show that activity classification accuracy is maintained above 80% within 32 second-cycle for sitting and 8 second-cycle for walking and standing. As a result, we can conclude optimal data collection cycle for activity recording is 8 seconds. This conclusion can be used the mobile health screening form.

Maintaining a proper lookout activity routine is integral to preventing ship collision accidents caused by human errors. Various subjective measures such as interviewing, self-report diaries, and questionnaires have been widely used to monitor the lookout activity patterns of navigators. An objective measurement of a lookout activity pattern classification system is required to improve lookout performance evaluation in a real navigation setting. The purpose of this study was to develop an objective navigator lookout activity classification system using wearable accelerometers. In the training session, 90.4% accuracy was achieved in classifying five fundamental lookout activities. The developed model was then applied to predict real-lookout activity in the second session during an actual ship voyage. 86.9% agreement was attained between the directly observed activity and predicted activity. Based on these promising results, the proposed unobstructed wearable system is expected to objectively evaluate navigator lookout patterns to provide a better understanding of lookout performance.

Travel trends are changing due to the prolonged COVID-19 pandemic and vaccine development. The analysis of pre and post Covid-19 tourism trends, according to a survey by Jeju Tourism Organization, shows that the search volume for overseas travel has decreased compared to 2018 and 2019, but search volume for Jeju travel and the number of tourists visiting Jeju Island Increased. In the case of tourists in Jeju, many use vehicles, mostly rental cars, for transportation due to the geographical characteristics of the island, and the number of electric vehicles is increasing in Jeju Island’s rental car services due to the strengthening of electric car policies. However, most of the existing research on tourists have been conducted using public data. Therefore, based on the means of transportation mainly used by tourists, electric vehicle driving data recorded for three years provided by Korea Electric Power Corporation Knowledge Data Network (KEPCO KDN) was classified into a total of 11 areas by weather and time requirements and classified through an artificial intelligence-based multiclassification model. In this study, tourist activity patterns were classified according to season, time zone, and climate conditions, but in the future, it can be used for recommendations and advertisements for tourist destinations by subdividing zones and adding information on users.

This paper reviews an implementation strategy of activity breakdown for the assessment of process time. In addition, the study proposes the classification models for estimating the process time of Time-Driven Activity-Based Costing (TDABC) based on various types of activity breakdown structures, including activity interface perspective, activity decomposition perspective and activity priority perspective.

최근 스마트폰의 보급 및 웹 서비스의 도입으로 온라인 사용자들은 대규모의 콘텐츠를 시간과 장소에 관계없이 접할 수 있게 되었다. 그러나 사용자들은 대규모의 콘텐츠 사이에서 원하는 콘텐츠를 찾는 데 어려움을 겪게 되었다. 이러한 문제를 해결하기 위해 다양한 분야에서 사용자 모델링 및 추천 시스템에 대한 연구가 활발하게 수행되었다. 그러나 정보 환경의 변화에 따른 시니어 계층의 적극적인 변화에도 불구하고 시니어 계층에 초점을 맞춘 사용자 모델링 및 추천 시스템에 대한 연구는 매우 부족한 실정이다. 이에 본 논문에서는 기계 학습 방법을 기반으로 스마트 시니어 계층의 선호도를 파악할 수 있는 모델링 방법을 제안하고, 스마트 시니어 분류 모델을 개발 한다. 이 결과, 스마트 시니어 계층의 선호도를 파악할 수 있을 뿐만 아니라 스마트 시니어 분류 모델 개발을 통해 시니어 사용자에게 가장 적합한 활동 및 콘텐츠를 제공하는 콘텐츠 추천 연구에 대한 발판을 마련하였다.

With the recent spread of smartphones and the introduction of web services, online users can access large-scale content regardless of time or place. However, users have had trouble finding the content they wanted among large-scale content. To solve this problem, user modeling and content recommendation system have been actively studied in various fields. However, in spite of active changes in senior groups according to the changes in information environment, research on user modeling and content recommendation system focused on senior groups are insufficient. In this paper, we propose a method of modeling smart senior based on their preference, and further develop a smart senior classification model using machine learning methods. As a result, we can not only grasp the preferences of smart seniors, but also develop a smart senior classification model, which is the foundation for the research of a recommendation system which will provide the activities and contents most suitable for senior groups.

This study examined health habit according to physical activity levels and BMI classification in university students using a health habit(smoking, drinking, drug uptake, dietary life, exercise, and stress solution) survey. The questionnaire survey was conducted with 411 university students residing in G Metropolitan City to provide basic information for the maintenance and promotion of health. The following conclusions were drawn. 1. The total score of health habit according to physical activity levels was significantly different in all factors of smoking, drinking, drug uptake, dietary life, exercise, and stress solution. Low-intensity exercise showed a high response in smoking, drinking, and drug uptake. Moderate-intensity exercise showed a high response in dietary life and exercise. High-intensity exercise showed a high response in stress solution. 2. The total score of health habit according to BMI classification was significantly different in all factors of smoking, drinking, drug uptake, dietary life, exercise, and stress solution. Low weight showed a high response in drinking and drug uptake. Normal weight showed a high response in smoking, dietary life, and exercise. Obesity showed a high response in stress solution. 3. The cross tabulation analysis of health habit by grade according to physical activity levels was significantly different in all factors of smoking, drinking, drug uptake, dietary life, exercise, and stress solution. The excellent grade of smoking, drinking, and drug uptake was high in low-intensity exercise. The excellent grade of dietary life was high in moderate- intensity exercise. The excellent grade of exercise and stress solution was high in high-intensity exercise. 4. The cross tabulation analysis of health habit by grade according to BMI classification was significantly different in all factors of smoking, drinking, drug uptake, dietary life, exercise, and stress solution. The excellent grade of smoking, drinking, drug uptake, dietary life, exercise, and stress solution was high in normal weight. The excellent grade of drinking and drug uptake was high in over-weight.

본 연구의 목적은 인공지능의 딥러닝을 활용하여 소셜미디어에서 공유되는 도시공원 이용자 활동사진을 분류하는 기초 모델을 만드는 것이다. 소셜미디어 데이터는 네이버 검색을 통해 수집된 도시공원 관련 사진들을 수집하여 분류모델에 활용하였다. 도시공원 특성 평가에 활용할 수 있는 지표인 자연성(naturalness), 잠재적 매력성(potential attraction), 활동(activity)을 기반으로 최종 21개의 분류 항목체계를 만들고, 항목별로 네이버에서 공유되는 실제 도시공원 사진을 수집하여 주석이 달린 데이터 세트를 구축했다. 수집한 사진 데이터 세트에 대해 커스텀(cuntom) CNN 모델과 사전 훈련된 CNN의 전이학습 모델을 설계하고 분석하였다. 연구결과, 가장 우수한 성능을 보였던 Xception 전이학습 모델이 최종적으로 도시공원 이용자 활동 이미지 분류모델로 선정되었으며, 그 외 다양한 평가 지표를 통해 모델을 평가했다. 본 연구는 소셜미디어에 공유되는 이용자 사진을 활용하여 도시공원 특성을 평가할 수 있는 지표로서 AI를 구축한 것에 의의가 있다. 딥러닝을 활용한 분류모델은 수동분류에 대한 한계를 보완하고, 대량의 도시공원 사진을 효율적으로 분류할 수 있어서 향후 도시공원의 모니터링 및 관리에 활용할 수 있는 유용한 방법이라고 할 수 있다.

This study aims to create a basic model for classifying the activity photos that urban park users shared on social media using Deep Learning through Artificial Intelligence. Regarding the social media data, photos related to urban parks were collected through a Naver search, were collected, and used for the classification model. Based on the indicators of Naturalness, Potential Attraction, and Activity, which can be used to evaluate the characteristics of urban parks, 21 classification categories were created. Urban park photos shared on Naver were collected by category, and annotated datasets were created. A custom CNN model and a transfer learning model utilizing a CNN pre-trained on the collected photo datasets were designed and subsequently analyzed. As a result of the study, the Xception transfer learning model, which demonstrated the best performance, was selected as the urban park user activity image classification model and evaluated through several evaluation indicators. This study is meaningful in that it has built AI as an index that can evaluate the characteristics of urban parks by using user-shared photos on social media. The classification model using Deep Learning mitigates the limitations of manual classification, and it can efficiently classify large amounts of urban park photos. So, it can be said to be a useful method that can be used for the monitoring and management of city parks in the future.

본 연구는 고령층의 경제활동상태에 근거한 경력경로를 유형화하고 각 유형의 결정요인을 분석하였다. 고령화패널조사 1~6차년도 조사자료를 활용하였고 1차년도 조사당시 55~64세이고 경제활동인구인 고 령층을 대상으로 시퀀스분석을 하여 유형화한 후 결정요인을 분석하였다. 분석결과 고령층의 경력경로는 이직형, 은퇴형, 자영업은퇴형, 임금근로유지형, 자영업유지형, 임금근로은퇴형, 시간제이직형, 농림어업유 지형의 8개 유형으로 구분할 수 있었다. 현재 자영업자이거나 농림어업 종사자인 경우 향후에도 동일한 경제활동상태를 유지할 가능성이 높았고 임금근로자인 경우 자영업유지형이나 농림어업유지형으로 분류 되는 경우는 극히 소수에 불과하여 임금근로자가 현재 일자리에서 퇴직하여 자영업이나 농림어업으로 전 직하여 성공적으로 정착하기는 매우 어렵다는 것을 보여주었다. 이러한 고령노동시장의 모습은 종사상 지위 간 이직이 활발하지 않고 이직을 한다 하더라도 대부분 실패하고 비경활인구됨을 보여주고 있었다. 한편 임금근로자의 경우 현재일자리의 시간당임금이 높을수록 향후 은퇴형이 될 가능성이 높은 것으로 나타나 생산성이 높은 임금근로자일수록 현재일자리에서 퇴직 후에 새로운 일자리를 찾기보다는 은퇴하 여 여가를 즐기는 것을 택함을 알 수 있었다. 이는 고령화사회에 대비하기 위해 이들의 생산성을 활용하 기 위한 재취업지원서비스 구축의 필요성을 시사하였다.

This study categorized the career path based on the economic status of the elderly and analyzed the determinants of each type. Survey data from the 1st to 6th waves of the KLoSA were used, and analyzed on the elderly who were 55 to 64 years old at the time of the 1st survey and were economically active population. As a result of the analysis, career paths of the elderly could be classified into 8 types. Those who are self-employed or those who are engaged in agriculture, forestry and fisheries are highly likely to maintain the same economic activity in the future. In the case of wage workers, only a few cases are classified as self-employment maintenance type or agriculture, forestry and fishery maintenance type. Such an image of the aged labor market showed that job change between job classes wasn’t active, and even if they did, most of them failed and were unemployed. On the other hand, the higher the hourly wage for the current job, the higher the possibility of retirement in the future. It implies that the more productive wage workers, the more they choose to retire and enjoy leisure rather than finding a new job.

An enhanced and optimized adaptive filter with optimal filter coefficients selection is proposed and implemented to resolve motion artifact issue in wearable ECG Recording. A two-electrode small size chest belt ECG system mounted with 3-axis accelerometer is implemented for ECG and activity ubiquitous recording in daily life. In ubiquitous ECG recording, ECG signal is often distorted due to different state of activity. Body movement incurs activity noise in ubiquitous ECG recording, and causing low accuracy in R-peak (heart beat) detection. Thus, a new adaptive filter methodology is proposed in this paper to look for an optimized filter coefficients base on different state of activity. A simple fuzzy rule-based algorithm is suggested for activity state classification and a set of high pass filter coefficient is applied base on different state of activity. In the case of low activity state, low high pass filter coefficient is used, whereas, in the case of high activity state, a high pass filter coefficient is used. The experiment result shows significant improvement of R-peak detection accuracy during fast movement activity state.

Bed Activity Monitoring System (BAMS) monitors and assess the mobility of people on a bed. This is a useful and critical application for patients with mobility issues after stroke or traumatic brain injury. The system is based on processing of data collected from a piezoelectric pressure sensor for discriminating mobility patterns. There are four different types of motion that were simulated by non-patient volunteers and data was collected. In this paper, two methods were used to extract feature parameters (autoregressive and cepstral coefficients) from the acquired data. Two classification algorithms, Euclidean Distance Measure (EDM) and Weighted Distance Measure (WDM) were used to classify and discriminate the mobility patterns of normal person (healthy subject) from people with mobility issues (patient subjects). Experimental result shows that the recognition rate using cepstral parameters was more effective compare to autoregressive parameters.

Purpose: The purpose of this study was to analyze of PCAs' activity classification system and time of PCAs who worked in Wamco. Method: The data were collected from 2 WAMCO and 308 subjects between February and August, 2007, by questionnaire and 24 time survey. The data were processed with SPSS Win 12.0. Result: In activity analysis, PCAs' activities were classified into 20 domains and 76 activities, which were hygiene, bathing, feeding & nutrition, elimination, respiration, skin care, exercise & transfer, problematic behavior control, communication, observation & measurement․comfort, medication, assisting test & treatment, reporting, environment management, patient belongings care, education attendance, indirect caregiving weekly/monthly PCAs' activity. And the PCAs' time analysis showed the average of 24hrs PCAs' time were 798.8 minutes, in which 46.8% were used in day-duty, 33.6% in evening-duty, and 19.6% in night-duty. There were no statistically significant difference in total PCAs time according to the type of industrial accidents and PCAs' type and qualification. But there were statistically significant difference in total PCAs time according to the type of PCAs (day-duty/all-night vigil. Conclusion: The results of this study can be utilized usefully and reasonally in deciding of PCAs staffing and PCAs' type and grade in WAMCO.

We report on the optical spectroscopic analysis of a Low Luminosity Quasi Stellar Objects (LLQSOs) sample at $z{\leq}0.06$ based on the Hamburg/ESO QSO survey (HES). To better relate the low-redshift Active Galactic Nucleus (AGN) to the QSO population it is important to study samples of the latter type at a level of detail similar to that of the low-redshift AGN. Powerful QSOs, however, are absent at low redshifts due to evolutionary effects and their small space density. Our understanding of the (distant) QSO population is, therefore, significantly limited by angular resolution and sensitivity. The LLQSOs presented here offer the possibility to study the faint end of this population at smaller cosmological distances and, therefore, in greater detail. This, in turn, provides information about the key ingredients with respect to fueling and feedback of QSOs, and their relative importance/strength. Here, we present results of the analysis of visible wavelength spectroscopy provided by the HES and the 6 Degree Field Galaxy Survey (6dFGS). Interesting differences in the taxonomy of the sources having both types of spectra have been noticed and will be discussed.

The discrimination of the source for xenon gases' release can provide an important clue for detecting the nuclear activities in the neighboring countries. In this paper, three machine learning techniques, which are logistic regression, support vector machine (SVM), and k-nearest neighbors (KNN), were applied to develop the predictive models for discriminating the source for xenon gases' release based on the xenon isotopic activity ratio data which were generated using the depletion codes, i.e., ORIGEN in SCALE 6.2 and Serpent, for the probable sources. The considered sources for the neighboring countries of South Korea include PWRs, CANDUs, IRT-2000, Yongbyun 5 MWe reactor, and nuclear tests with plutonium and uranium. The results of the analysis showed that the overall prediction accuracies of models with SVM and KNN using six inputs, all exceeded 90%. Particularly, the models based on SVM and KNN that used six or three xenon isotope activity ratios with three classification categories, namely reactor, plutonium bomb, and uranium bomb, had accuracy levels greater than 88%. The prediction performances demonstrate the applicability of machine learning algorithms to predict nuclear threat using ratios of xenon isotopic activity.

Classification of human thought is an emerging research field that may allow us to understand human brain functions and further develop advanced brain-computer interface (BCI) systems. In the present study, we introduce a new approach to classify various mental states from noninvasive electrophysiological recordings of human brain activity. We utilized the full spatial and spectral information contained in the electroencephalography (EEG) signals recorded while a subject is performing a specific mental task. For this, the EEG data were converted into a 2D spatiospectral pattern map, of which each element was filled with 1, 0, and -1 reflecting the degrees of event-related synchronization (ERS) and event-related desynchronization (ERD). We evaluated the similarity between a current (input) 2D pattern map and the template pattern maps (database), by taking the inner-product of pattern matrices. Then, the current 2D pattern map was assigned to a class that demonstrated the highest similarity value. For the verification of our approach, eight participants took part in the present study; their EEG data were recorded while they performed four different cognitive imagery tasks. Consistent ERS/ERD patterns were observed more frequently between trials in the same class than those in different classes, indicating that these spatiospectral pattern maps could be used to classify different mental states. The classification accuracy was evaluated for each participant from both the proposed approach and a conventional mental state classification method based on the inter-hemispheric spectral power asymmetry, using the leave-one-out cross-validation (LOOCV). An average accuracy of 68.13% (${\pm}9.64%$) was attained for the proposed method; whereas an average accuracy of 57% (${\pm}5.68%$) was attained for the conventional method (significance was assessed by the one-tail paired $t$-test, $p$ < 0.01), showing that the proposed simple classification approach might be one of the promising methods in discriminating various mental states.

Secondary metabolite-based chemotaxonomic classification of Streptomyces (8 species, 14 strains) was performed using ultraperformance liquid chromatography-quadrupole-time-offlight-mass spectrometry with multivariate statistical analysis. Most strains were generally well separated by grouping under each species. In particular, S. rimosus was discriminated from the remaining sevens pecies (S. coelicolor, S. griseus, S. indigoferus, S. peucetius, S. rubrolavendulae, S. scabiei, and S. virginiae) in partial least squares discriminant analysis, and oxytetracycline and rimocidin were identified as S. rimosus-specific metabolites. S. rimosus also showed high antibacterial activity against Xanthomonas oryzae pv. oryzae, the pathogen responsible for rice bacterial blight. This study demonstrated that metabolite-based chemotaxonomic classification is an effective tool for distinguishing Streptomyces spp. and for determining their species-specific metabolites.

The use of the classification and regression tree (CART) methodology was studied in a quantitative structure-activity relationship (QSAR) context on a data set consisting of the binding affinities of 39 imidazobenzodiazepines for the α1 benzodiazepine receptor. The 3-D structures of these compounds were optimized using HyperChem software with semiempirical AM1 optimization method. After optimization a set of 1481 zero-to three-dimentional descriptors was calculated for each molecule in the data set. The response (dependent variable) in the tree model consisted of the binding affinities of drugs. Three descriptors (two topological and one 3D-Morse descriptors) were applied in the final tree structure to describe the binding affinities. The mean relative error percent for the data set is 3.20%, compared with a previous model with mean relative error percent of 6.63%. To evaluate the predictive power of CART cross validation method was also performed.

This work aimed to classify Aspergillus (8 species, 28 strains) by using a secondary metabolite profile-based chemotaxonomic classification technique. Secondary metabolites were analyzed by liquid chromatography ion-trap mass spectrometry (LC-IT-MS) and multivariate statistical analysis. Most strains were generally well separated from each section. A. lentulus was discriminated from the other seven species (A. fumigatus, A. fennelliae, A. niger, A. kawachii, A. flavus, A. oryzae, and A. sojae) with partial least-squares discriminate analysis (PLS-DA) with five discriminate metabolites, including 4,6-dihydroxymellein, fumigatin, 5,8-dihydroxy-9-octadecenoic acid, cyclopiazonic acid, and neosartorin. Among them, neosartorin was identified as an A. lentulus-specific compound that showed anticancer activity, as well as antibacterial effects on Staphylococcus epidermidis. This study showed that metabolite-based chemotaxonomic classification is an effective tool for the classification of Aspergillus spp. with species-specific activity.

소아청소년은 성인과는 다른 신체적, 생리적 특성이 있으므로 동일한 신체활동을 할지라도 에너지소비량이 다름이 보고된 바 있다. 그러나 지금까지 국내에서 수행된 소아청소년 대상 연구에서 이들의 에너지소비량을 평가 시, 성인 대상으로 측정한 에너지소비량 또는 성인대상 18단계 신체활동분류표가 이용되어왔다. 최근 소아청소년을 대상으로 한 미국의 자료와 국내의 일부 자료를 토대로 4단계 연령대 (6-9세, 10-12세, 13-15세 및 16-18세)에 따라 서로 다른 에너지당량 (METs)을 제시하는 한국 소아청소년을 위한 신체활동분류표가 보고되었다. 이에 본 연구의 1부에서 에너지소비량 측정방법의 gold standard로 알려져 있는 이중표식수법을 이용하여 소아청소년을 위한 신체활동분류표의 타당도를 평가한 결과, 이의 활용 가능성을 확인하였다. 2부에서는 세계보건기구에서 제안한 방법 (TEE = REE × PAL), 즉 간접열량계로 측정한 REE과 소아청소년 신체활동분류표로 측정한 PAL을 이용하여 166명의 소아청소년 (초·중·고등학생)의 일일 총에너지소비량을 산출하였다. 이를 기준으로 한국인 영양소 섭취기준에서 제시한 공식을 이용한 EER을 비교한 결과, 과대평가 비율이 47.3%로 나타났다. 또한 소아청소년 신체활동분류표를 이용하여 산출된 신체활동 강도별 및 신체활동 유형별 소비시간이 성별 및 연령대별로 차이가 있음을 알 수 있었다. 앞으로 중학생 및 고등학생을 포함하는 다수의 소아청소년을 대상으로 이중표식수법을 이용하여 일일 총에너지소비량을 측정하는 연구가 수행되어, 소아청소년을 위한 신체활동분류표의 타당도가 폭넓게 평가되어야 한다. 또한 이와 같은 이중표식수법 연구를 통하여 우리나라 소아청소년을 위한 에너지필요추정량 산출 공식의 타당도 평가와 함께 한국의 소아청소년을 위한 에너지필요추정량공식이 개발되어야 할 것이다. 또한 학업부담이 많은 고등학생 뿐만 아니라, 중학생 및 초등학생에서도 강도있는 신체활동의 소비시간이 매우 낮은 것으로 나타났으므로 우리나라 소아청소년의 건강 및 비만의 예방 및 관리를 위하여 이들의 에너지소비량을 증진시킬 수 있는 프로그램의 개발 및 적용이 요구된다

Purpose: The purpose of the first part of this study was to evaluate the validity of the physical activity classification table for youth (Youth-PACT). The second part of this study was aimed at comparing the estimated energy requirement (EER) with the total energy expenditure (TEE) and evaluating the physical activity patterns of Korean children and adolescents. Methods: The subjects of the first part of the study were 17 children aged 10 to 12 years, and their total energy expenditure (TEE_DLW) was measured using the double labeled water (DLW) method. A total of 166 children and adolescents aged 6-18 years participated in the second part of this study. Their resting energy expenditure (REE) was measured using indirect calorimetry and the TEE_Youth-PACT and physical activity level were calculated by applying the Youth-PACT to the physical activity diary prepared by the subjects. Results: In the first part of this study, there were no significant differences between the TEE_DLW and the TEE_Youth-PACT. The TEE_Youth-PACT accurately predicted TEE_DLW in 37.5% of the subjects. In the second part of the study, the rates at which EER accurately predicted TEE _YouthPACT and overestimated TEE _Youth-PACT were 29.6% and 47.3%, respectively. The time spent based on intensity of physical activity and the physical activity categories which were obtained using Youth-PACT showed different patterns according to sex and age group. Age showed significant positive correlations with REE, TEE, and the time spent in sedentary behavior, but age was significantly negatively correlated with REE/body weight, TEE/body weight, and the time spent in low-intensity and high-intensity activities. Conclusion: The results of this study showed that the Youth-PACT can be used to evaluate the TEE and PAL of children and adolescents. However, further studies are needed to validate the TEE_Youth-PACT and to set the EER for children and adolescents.

최근 컴퓨터를 이용한 다양한 인식 문제를 해결하기 위해 딥 러닝을 적용하는 사례가 늘어나고 있다. 딥 러닝은 학습에 필요한 요소를 학습데이터를 통해 스스로 도출해내기 때문에, 수작업(hand-craft)을 통해 특징을 도출하던 기존의 기계학습 방법보다 더 많은 장점을 갖는다. 행동인식을 위한 기존의 심층 신경망은 비디오 데이터를 일정 프레임의 이미지로 분할한 후, 분할된 각 이미지 사이의 시간적 연계성 분석을 통해 행동을 분류한다. 그러나 이러한 신경망은 소규모 행동 클래스를 갖는 분류 문제에서 학습 데이터의 부족 문제 및 과적합(overfitting) 문제로 인해 이를 실제 문제에 적용하기 어려운 경우가 많다. 이에 본 논문에서는 5가지의 소규모 행동 클래스를 정의하고, 기존 행동 인식 신경망의 최적화를 통해 이를 분류하였다. 700개의 비디오데이터를 통해 행동 데이터베이스를 구성하였고, 약 74.00%의 분류 정확도를 얻을 수 있었다.

Recently, Deep learning has been used successfully to solve many recognition problems. It has many advantages over existing machine learning methods that extract feature points through hand-crafting. Deep neural networks for human activity recognition split video data into frame images, and then classify activities by analysing the connectivity of frame images according to the time. But it is difficult to apply to actual problems which has small-scale activity classes. Because this situations has a problem of overfitting and insufficient training data. In this paper, we defined 5 type of small-scale human activities, and classified them. We construct video database using 700 video clips, and obtained a classifying accuracy of 74.00%.

이 연구의 목적은 초등학교 6학년 학생의 분류활동에서 나타나는 사고 유형, 과정과 특징을 분석하는 것이다. 이러한 목적을 달성하기 위하여 분류활동 수행에 적합한 2가지 도구를 개발하였다. 첫 번째는 속성이 분명하게 드러나는 인공자극카드이고, 두 번째는 속성이 잘 드러나지 않는 자연자극카드이다. 서울시 영등포구 소재 D초등학교 6학년 8명을 대상으로 질적 연구를 수행하였다. 자료는 피험자의 과제 수행과정을 녹화한 비디오테이프, 피험자의 분류 기록지, 연구자의 피험자 행동 관찰, 피험자와의 면담 등 자료 삼각측정법에 의해 획득하였다. 연구결과, 6학년 학생들은 분류활동에서 속성 관찰, 속성 평가, 예비 점검, 기준 선택, 샘플 동정의 다섯 가지 유형의 사고를 하였으며, 모든 항목을 분류할 때까지 속성 관찰 $\rightarrow$ 속성 평가 $\rightarrow$ 예비 점검 $\rightarrow$ 기준 선택 $\rightarrow$ 샘플 동정의 과정을 반복하였다. 그리고 인지 경제성을 활용하여 분류하여 분류하였다. 이상의 연구 결과는 과학 분류 학습 지도에 다음과 같은 시사점을 줄 수 있다. 첫째,교사가 학생들의 분류 사고과정을 이해한다면, 보다 효과적인 분류학습 지도가 가능할 것이다. 둘째, 분류사고 과정의 각 단계를 고려한 단계별 학습지도가 필요하다.

The purpose of this study was to investigate 6th grade pupil's thoughts during classification activities. Two suitable tools in classification activity achievement were developed to achieve this purpose. The first was an artificial stimulus card in which the attribute was prominent; and the other a natural stimulus card in which the attribute was less prominent. Participants of the study were 8 6th grade pupils from D elementary school in Yeongdeungpo-gu, Seoul. Data were collected from interviews with the pupils, the pupil's recordings of classification, the investigator's observation of pupil's actions, and video recordings of the pupil's subject classification process. Results found in this study were as following. First, when doing classification 6th grade pupils considered attribute observation, attribute estimation, preliminary inspection, criteria selection, and sample identification. Second, 6th grade pupil classification thought process was found to be repetitive, passing through the steps of attribute observation, attribute estimation, preliminary inspection, criteria selection, and lastly, sample identification. Third, 6th grade pupils took advantage of cognitive economic efficiency. Study findings also revealed guidance for the teaching and learning of scientific classification. First, once teachers understand the classification thought process of students, more effective classification guidance will be possible. Second, it is necessary that guidance fit each step of the classification thought process.