• Geomechanics and Engineering
• /
• 제25권4호
• /
• pp.317-330
• /
• 2021

This paper reports the results of advanced data analysis involving artificial neural networks for the prediction of the unconfined compressive strength of granite using only two non-destructive test indexes. A data-independent site-independent unbiased database comprising 182 datasets from non-destructive tests reported in the literature was compiled and used to train and develop artificial neural networks for the prediction of the unconfined compressive strength of granite. The results show that the optimum artificial network developed in this research predicts the unconfined compressive strength of weak to very strong granites (20.3-198.15 MPa) with less than ±20% deviation from the experimental data for 70% of the specimen and significantly outperforms a number of available models available in the literature. The results also raise interesting questions with regards to the suitability of the Pearson correlation coefficient in assessing the prediction accuracy of models.

• 한국환경복원기술학회지
• /
• 제24권5호
• /
• pp.25-34
• /
• 2021

Increase in avian species richness is one of the important issues of urban biodiversity policies, since it can promote diverse ecosystem services such as seed dispersal, education, and pollination. However, though human disturbance can significantly affect avian species richness, there are limited studies on the way to reflect the dynamics of floating population. Therefore, this study analyzed the spatial relationship between avian species richness, floating population, and vegetation cover using telecommunications information to identify the areas that requiring targeted monitoring and restoration action. Bivariate Local Moran's I was applied to identify LISA cluster map that showing representative biotopes, which reflect significant spatial relationship between species richness and population distribution. Edge density and distribution of ndvi were identified for evaluating relative adequacy of selected biotopes to strengthen the robust biodiversity network. This study offers insight to consider human disturbance in spatial context using innovative big data to increase the effectiveness of urban biodiversity measures.

• 로봇학회논문지
• /
• 제18권1호
• /
• pp.122-126
• /
• 2023

An unexpected occlusion in a real life, not in a laboratory, can be more fatal to neural networks than expected. In addition, it is virtually impossible to create a network that learns all the environmental changes as well as occlusions. Therefore, we propose an alternative approach in which the architecture and number of parameters remain unchanged while adapting to occlusion circumstances. Learning method with the term Conversion Learning classifies them more robustly by converting the weights from various occlusion situations. The experiments on MNIST dataset showed a 3.07 [%p] performance improvement over the baseline CNN model in a situation where most objects are occluded and unknowing what occlusion will appear in advance. The experimental results suggest that Conversion Learning is an efficient method to respond to environmental changes such as occluded images.

• 생태와환경
• /
• 제50권4호
• /
• pp.478-482
• /
• 2017

In order to identify key nations and bird species of conservation concern we described multinational collaborations as defined using network analysis linked by birds that are found in all nations in the network. We used network analysis to assess the patterns in bird occurrence for 10,422 bird inventories from 244 countries and territories. Nations that are important in multinational collaborations for bird conservation were assessed using the centrality measures, closeness and betweenness centrality. Countries important for the multinational collaboration of bird conservation were examined based on their centrality measures, which included closeness and betweenness centralities. Comparatively, the co-occurrence network was divided into four groups that reveal different biogeographical structures. A group with higher closeness centrality included countries in southern Africa and had the potential to affect species in many other countries. Birds in countries in Asia, Australia and the South Pacific that are important to the cohesiveness of the global network had a higher score of betweenness centrality. Countries that had higher numbers of bird species and more extensively distributed bird species had higher centrality scores; in these countries, birds may act as excellent indicators of trends in the co-occurrence bird network. For effective bird conservation in the world, much stronger coordination among countries is required. Bird co-occurrence patterns can provide a suitable and powerful framework for understanding the complexity of co-occurrence patterns and consequences for multinational collaborations on bird conservation.

• 한국보건간호학회지
• /
• 제30권2호
• /
• pp.311-325
• /
• 2016

Purpose: This study was conducted to examine the individual and community level factors associated with physical activity and to identify its relative effects using a multilevel analysis among Korean adults in certain counties. Methods: A cross-sectional data of 39,547 adults (age range of 19~64 years) living in 82 counties from the 2013 Korean Community Health Survey (KCHS) was analyzed. Individual and social correlates from KCHS and physical environmental data from the Korean Statistical Information Service were collected. A multilevel logistic regression was performed using Stata 10.0 IC. Results: Multilevel analyses showed that the effect of social and physical environmental on engaging in moderate or vigorous physical activity (MVPA) was significant in comparison to the influence of individual correlates. The individual factors that were associated with participating in MVPA included gender, marital status, education, job, and household income. In the community level, social environmental factors associated with engagement in MVPA were higher satisfaction with healthcare service (OR=3.410, 95% CI=1.109~11.269), a high level of social support (OR=5.920, 95% CI=1.459~22.657) and social network (OR=1.025, 95% CI= 1.017~1.032). Conclusion: To promote moderate or vigorous physical activity in Korean adults in some counties, social environmental factors should be considered along with individual correlates.

• International Journal of Computer Science & Network Security
• /
• 제21권7호
• /
• pp.87-92
• /
• 2021

Analysis of current state of construction industry functioning in Ukraine allows us to identify a number of problems having negative impact on sustainable development of construction industry, especially in terms of its organization. Therefore, it is absolutely essential to study existing methods of organization system supplying construction sites with necessary material resources. Companies can develop their own logistics departments, which independently solve logistics issues related to transportation organization and management, accounting and inventory management, acquisition and warehousing, intercommunication (ability to obtain both final and intermediate information during transporting materials). Using a complex of methods is substantiated: the hierarchy analysis method (Saati's method), the network method, the defect elimination algorithm DEA, the transportation problem that finds optimal problem solutions for construction sector with the purpose of rational supplying uninterrupted construction with building resources in the designed model "provider-transportation-costs".

• Steel and Composite Structures
• /
• 제42권6호
• /
• pp.733-745
• /
• 2022

This study proposed a robust artificial intelligence (AI) model based on the social behaviour of the imperialist competitive algorithm (ICA) and artificial neural network (ANN) for modelling the deflection of reinforced concrete beams, abbreviated as ICA-ANN model. Accordingly, the ICA was used to adjust and optimize the parameters of an ANN model (i.e., weights and biases) aiming to improve the accuracy of the ANN model in modelling the deflection reinforced concrete beams. A total of 120 experimental datasets of reinforced concrete beams were employed for this aim. Therein, applied load, tensile reinforcement strength and the reinforcement percentage were used to simulate the deflection of reinforced concrete beams. Besides, five other AI models, such as ANN, SVM (support vector machine), GLMNET (lasso and elastic-net regularized generalized linear models), CART (classification and regression tree) and KNN (k-nearest neighbours), were also used for the comprehensive assessment of the proposed model (i.e., ICA-ANN). The comparison of the derived results with the experimental findings demonstrates that among the developed models the ICA-ANN model is that can approximate the reinforced concrete beams deflection in a more reliable and robust manner.

• 한국정보통신학회:학술대회논문집
• /
• 한국해양정보통신학회 2011년도 추계학술대회
• /
• pp.839-842
• /
• 2011

최근 무선 센서노드의 에너지 제약을 해결하기 위해 에너지 하비스팅 기반 무선 센서 네트워크에 대한 연구가 활발히 진행되고 있다. 에너지 하비스팅 기반 무선 센서 네트워크는 태양열, 풍력, 지력 에너지 등의 환경 에너지를 센서 노드의 전력으로 사용하여 노드의 생존 시간을 향상시키는 기술이다. 이러한 에너지 하비스팅 환경에서 기존의 에너지가 제약된 환경을 배경으로 개발된 라우팅 프로토콜을 사용할 경우, 하비스팅 장치로부터 축적되는 에너지가 경로 설정 시 반영되지 않는 문제점이 발생한다. 또한 분산되지 않은 경로들은 네트워크의 수명을 단축하게 된다. 따라서 본 논문에서는 각 노드간의 경로를 분산시키는 알고리즘을 제안한다. 시뮬레이터를 통해 제안하는 알고리즘을 사용했을 경우 노드의 경로가 다양하게 반영되는 것을 보여주었다.

• 한국경제지리학회지
• /
• 제15권3호
• /
• pp.437-455
• /
• 2012

전 세계적으로 환경문제가 큰 이슈가 됨에 따라, 우리나라에서도 신재생에너지에 대한 관심의 일환으로 전라도를 중심으로 많은 태양광발전소가 건설되었다. 그러나 입지 선정 과정에서 환경 사회적 영향을 충분히 고려하지 않아 각종 문제점들이 야기되었으며, 기존 시설의 입지 타당성 평가와 지속가능한 입지 모색이 요구되었다. 현재까지 환경을 고려한 태양광발전소의 입지 선정 연구는 미미하며, 기존 시설을 평가한 연구도 진행된 바 없다. 따라서 본 연구에서는 전라도에서 환경 사회적 영향을 최소화하는 지속가능한 입지를 제안하고, 현존하는 시설의 입지 타당성을 평가하였다. 이를 위해 연구방법으로 크리깅(Kriging)과 네트워크분석법(Analytic Network Process; ANP)를 이용하였다. 네트워크분석법을 통해 도출된 가중치로 모든 공간데이터를 통합하여 최적의 입지를 도출하고, 환경 사회적으로 영향을 줄 수 있는 구역을 최적지에서 제외하였다. 그 결과, 전반적으로 서해안과 남해안 부근에 적합지가 많은 것으로 나타났다. 또한, 환경 사회적 영향을 고려한 최종 입지 적합도와 기존 시설의 위치를 비교 분석한 결과 총 81곳 중 23곳이 부적합한 태양광발전소로 판단되었다. 본 연구의 결과를 통해 사회적 갈등 및 환경문제를 방지하는 지속가능한 입지 선정이 가능할 것이며, 기존 입지를 객관적으로 파악하는 지표를 제공함으로써 대책 마련에 도움을 줄 것으로 기대된다.

• 한국항만경제학회지
• /
• 제39권2호
• /
• pp.15-23
• /
• 2023

최근 ESG 경영의 필요성, 중요성이 증대되는 상황에서 수출입 물류의 거점인 컨테이너터미널에 대한 ESG관련 중요 변수를 ESG 평가기관의 다양한 변수 중에서 선정하는 것은 컨테이너터미널의 ESG 경영 전략을 수립하는데 도움이 될 것이라고 판단되어 이번 연구를 진행하게 되었다. 본 연구의 목적을 효과적으로 달성하기 위하여 문헌연구, 인터뷰를 바탕으로 KCGS ESG 모범규준에서 ESG관련 항목을 31개를 추출하고 변수의 축소 및 그룹핑을 위하여 빈도수 분석을 실시하고 Gephi 0.9 프로그램을 활용한 워드네트워크분석을 통해 변수의 연결도와 중심성을 분석하여 변수를 선정하였다. 본 연구의 실증분석의 결과를 요약하면 다음과 같다. 첫째, KCGS ESG 변수의 31개 항목의 빈도수 분석의 결과 친환경경영(E) 요소는 친환경 시설설비투자(22), 환경인식교육(13), 실무조직구성(7), 환경인증(6)의 순으로 빈도수가 높았다. 사회적책임경영(S) 요소는 근로자인권보호 및 지역사회 기여(21), 안전교육(20), 안전인증획득(17), 감염병 대응(10)의 순으로 빈도수가 높았다. 지배구조 및 윤리경영(Governance) 요소는 부정부패 척결(17), 투명경영(14), 이해관계자 상호협조(11), 지속가능경영보고서(4) 순으로 빈도수가 높았다. 둘째, 워드네트워크분석결과 친환경경영(E) 요소의 평균가중차수인 Weighted Degree는 친환경 시설 설비투자(54), 환경인식교육(45), 실무조직구성(31), 환경인증(32)의 순서로 나왔다. 중심성, 연결성 지수인 Page Ranks 순서는 환경인식교육(0.0765), 임직원 참여유도(0.0765), 친환경 시설 설비 투자(0.0761), 실무조직구성(0.0761), 환경인증(0.0761)로 나왔다. 사회적 책임 경영(S) 요소의 평균가중차수인 Weighted Degree는 근로자 인권보호 및 지역사회기여(68), 안전교육(63), 안전인증획득(59), 감염병대응(40)의 순서로 나왔다. 중심성, 연결성 지수인 Page Ranks 순서는 근로자 인권보호 및 지역사회 기여(0.165), 안전교육(0.153), 안전인증획득(0.144), 감염병 대응(0.102)로 나왔다. 지배구조 및 윤리경영(G) 요소의 평균가중차수인 Weighted Degree는 부정부패척결(27), 투명경영(24), 이해관계자 상호협조(19), 지속가능경영보고서(9)의 순서로 나왔다. 중심성, 연결성 지수인 Page Ranks 순서는 부정부패척결(0.241), 투명경영(0.216), 이해관계자 상호협조(0.174), 이사의 역할과 책임(0.105), 주주권보호(0.097), 지속가능경영보고서(0.096)로 나왔다.