• 한국물환경학회지
• /
• 제38권6호
• /
• pp.292-305
• /
• 2022

The Yeongsan River, one of the four largest rivers in South Korea, has been facing difficulties with water quality management with respect to algal bloom. The algal bloom menace has become bigger, especially after the construction of two weirs in the mainstream of the Yeongsan River. Therefore, the prediction and factor analysis of Chlorophyll-a (Chl-a) concentration is needed for effective water quality management. In this study, Chl-a prediction model was developed, and the performance evaluated using machine and deep learning methods, such as Deep Neural Network (DNN), Random Forest (RF), and eXtreme Gradient Boosting (XGBoost). Moreover, the correlation analysis and the feature importance results were compared to identify the major factors affecting the concentration of Chl-a. All models showed high prediction performance with an R² value of 0.9 or higher. In particular, XGBoost showed the highest prediction accuracy of 0.95 in the test data.The results of feature importance suggested that Ammonia (NH₃-N) and Phosphate (PO₄-P) were common major factors for the three models to manage Chl-a concentration. From the results, it was confirmed that three machine learning methods, DNN, RF, and XGBoost are powerful methods for predicting water quality parameters. Also, the comparison between feature importance and correlation analysis would present a more accurate assessment of the important major factors.

• 국제학술발표논문집
• /
• The 8th International Conference on Construction Engineering and Project Management
• /
• pp.463-481
• /
• 2020

Cutter suction dredgers (CSDs) are widely used in various dredging constructions such as channel excavation, wharf construction, and reef construction. During a CSD construction, the main operation is to control the swing speed of cutter to keep the slurry concentration in a proper range. However, the slurry concentration cannot be monitored in real-time, i.e., there is a "time-lag effect" in the log of slurry concentration, making it difficult for operators to make the optimal decision on controlling. Concerning this issue, a solution scheme that using real-time monitored indicators to predict current slurry concentration is proposed in this research. The characteristics of the CSD monitoring data are first studied, and a set of preprocessing methods are presented. Then we put forward the concept of "index class" to select the important indices. Finally, an ensemble learning algorithm is set up to fit the relationship between the slurry concentration and the indices of the index classes. In the experiment, log data over seven days of a practical dredging construction is collected. For comparison, the Deep Neural Network (DNN), Long Short Time Memory (LSTM), Support Vector Machine (SVM), Random Forest (RF), Gradient Boosting Decision Tree (GBDT), and the Bayesian Ridge algorithm are tried. The results show that our method has the best performance with an R² of 0.886 and a mean square error (MSE) of 5.538. This research provides an effective way for real-time predicting the slurry concentration of CSDs and can help to improve the stationarity and production efficiency of dredging construction.

• Journal of Animal Science and Technology
• /
• 제65권5호
• /
• pp.971-988
• /
• 2023

This study evaluated the effects of supplementing solubles from shredded, steam-exploded pine particles (SSPP) on growth performances, plasma biochemicals, and microbial composition in broilers. The birds were reared for 28 days and fed basal diets with or without the inclusion of SSPP from 8 days old. There were a total of three dietary treatments supplemented with 0% (0% SSPP), 0.1% (0.1% SSPP) and 0.4% (0.4% SSPP) SSPP in basal diets. Supplementation of SSPP did not significantly affect growth or plasma biochemicals, but there was a clear indication of diet-induced microbial shifts. Beta-diversity analysis revealed SSPP supplementation-related clustering (ANOSIM: r = 0.31, p < 0.01), with an overall lower (PERMDISP: p < 0.05) individual dispersion in comparison to the control group. In addition, the proportions of the Bacteroides were increased, and the relative abundances of the families Vallitaleaceae, Defluviitaleaceae, Clostridiaceae, and the genera Butyricicoccus and Anaerofilum (p < 0.05) were significantly higher in the 0.4% SSPP group than in the control group. Furthermore, the linear discriminant analysis effect size (LEfSe) also showed that beneficial bacteria such as Ruminococcus albus and Butyricicoccus pullicaecorum were identified as microbial biomarkers of dietary SSPP inclusion (p < 0.05; | LDA effect size | > 2.0). Finally, network analysis showed that strong positive correlations were established among microbial species belonging to the class Clostridia, whereas Erysipelotrichia and Bacteroidia were mostly negatively correlated with Clostridia. Taken together, the results suggested that SSPP supplementation modulates the cecal microbial composition of broilers toward a "healthier" profile.

• IMMUNE NETWORK
• /
• 제20권6호
• /
• pp.49.1-49.15
• /
• 2020

C-C chemokine receptor type 5 (CCR5) regulates the trafficking of various immune cells to sites of infection. In this study, we showed that expression of CCR5 and its ligands was rapidly increased in the kidney after systemic Candida albicans infection, and infected CCR5^-/- mice exhibited increased mortality and morbidity, indicating that CCR5 contributes to an effective defense mechanism against systemic C. albicans infection. The susceptibility of CCR5^-/- mice to C. albicans infection was due to impaired fungal clearance, which in turn resulted in exacerbated renal inflammation and damage. CCR5-mediated recruitment of NK cells to the kidney in response to C. albicans infection was necessary for the anti-microbial activity of neutrophils, the main fungicidal effector cells. Mechanistically, C. albicans induced expression of IL-23 by CD11c⁺ dendritic cells (DCs). IL-23 in turn augmented the fungicidal activity of neutrophils through GM-CSF production by NK cells. As GM-CSF potentiated production of IL-23 in response to C. albicans, a positive feedback loop formed between NK cells and DCs seemed to function as an amplification point for host defense. Taken together, our results suggest that CCR5-mediated recruitment of NK cells to the site of fungal infection is an important step that underlies innate resistance to systemic C. albicans infection.

• IMMUNE NETWORK
• /
• 제24권3호
• /
• pp.19.1-19.15
• /
• 2024

The influenza virus poses a global health burden. Currently, an annual vaccine is used to reduce influenza virus-associated morbidity and mortality. Most influenza vaccines have been developed to elicit neutralizing Abs against influenza virus. These Abs primarily target immunodominant epitopes derived from hemagglutinin (HA) or neuraminidase (NA) of the influenza virus incorporated in vaccines. However, HA and NA are highly variable proteins that are prone to antigenic changes, which can reduce vaccine efficacy. Therefore, it is essential to develop universal vaccines that target immunodominant epitopes derived from conserved regions of the influenza virus, enabling cross-protection among different virus variants. The internal proteins of the influenza virus serve as ideal targets for universal vaccines. These internal proteins are presented by MHC class I molecules on Ag-presenting cells, such as dendritic cells, and recognized by CD8 T cells, which elicit CD8 T cell responses, reducing the likelihood of disease and influenza viral spread by inducing virus-infected cell apoptosis. In this review, we highlight the importance of CD8 T cell-mediated immunity against influenza viruses and that of viral epitopes for developing CD8 T cell-based influenza vaccines.

• Applied Microscopy
• /
• 제51권
• /
• pp.20.1-20.12
• /
• 2021

We explored the link between mitochondrial biogenesis and mitochondrial morphology using transmission electron microscopy (TEM) in lymphoblasts of pediatric acute lymphoblastic leukemia (ALL) patients and compared these characteristics between tumors and control samples. Gene expression of mitochondrial biogenesis markers was analysed in 23 ALL patients and 18 controls and TEM for morphology analysis was done in 15 ALL patients and 9 healthy controls. The area occupied by mitochondria per cell and the cristae cross-sectional area was observed to be significantly higher in patients than in controls (p-value=0.0468 and p-value<0.0001, respectively). The mtDNA copy numbers, TFAM, POLG, and c-myc gene expression were significantly higher in ALL patients than controls (all p-values<0.01). Gene Expression of PGC-1α was higher in tumor samples. The analysis of the correlation between PGC-1α expression and morphology parameters i.e., both M/C ratio and cristae cross-sectional area revealed a positive trend (r=0.3, p=0.1). The increased area occupied by mitochondria and increased cristae area support the occurrence of cristae remodelling in ALL. These changes might reflect alterations in cristae dynamics to support the metabolic state of the cells by forming a more condensed network. Ultrastructural imaging can be useful for affirming changes occurring at a subcellular organellar level.

• 대한원격탐사학회지
• /
• 제39권5_3호
• /
• pp.933-948
• /
• 2023

대기 중 에어로졸은 인체에 악영향을 끼칠 뿐 아니라 기후 시스템에도 직간접적인 영향을 미치므로 에어로졸의 특성과 시공간적인 분포에 대한 이해는 매우 중요하다. 이를 위해 위성기반 관측을 통해 에어로졸 광학 두께(Aerosol Optical Depth, AOD)를 산출하여 에어로졸을 모니터링하는 다양한 연구가 수행되어 왔다. 하지만 이는 주로 조견표를 활용한 역 산출 알고리즘에 기반하여 이루어지기 때문에 많은 계산량을 요구하며 불확실성이 존재한다. 따라서, 본 연구에서는 Geostationary Ocean Color Imager-II (GOCI-II)의 대기상한반사도와 30일 동안의 대기상한반사도 중 최솟값과 관측 시점 값의 차이 값, 수치 모델 기반 기상학적 변수 등을 활용하여 기계학습 기반 고해상도 AOD 직접 산출 알고리즘을 개발하였다. Light Gradient Boosting Machine (LGBM) 기법이 사용되었으며, 추정된 결과는 지상 관측 자료인 Aerosol Robotic Network (AERONET) AOD를 활용하여 랜덤, 시간 및 공간별 N-fold 교차검증을 통해 검증되었다. 세 가지 교차검증 결과 R²=0.70-0.80, RMSE=0.08-0.09, 기대오차(Expected Error, EE) 안에 있는 비율은 75.2-85.1% 수준으로 안정적인 성능을 보였다. Shapley Additive exPlanations (SHAP) 분석에서는 반사도 관련 변수들이 기여도의 상위권 대부분을 차지하고 있는 것을 통해 반사도 자료가 AOD 추정에 많은 기여를 하는 것을 확인하였다. 서울과 울산 지역에 대한 시간 별 AOD의 공간 분포를 분석한 결과, 개발된 LGBM 모델은 시간의 흐름에 따라 AERONET AOD 값과 유사한 수준으로 AOD를 추정하고 있었다. 이를 통해 높은 시공간 해상도(i.e., 시간별, 250 m)에서의 AOD 산출이 가능함을 확인하였다. 또한, 산출 커버리지 비교에서 LGBM 모델의 평균 산출 빈도가 GOCI-II L2 AOD 산출물 대비 8.8%가량 증가한 것을 통해 기존 물리모델기반 AOD 산출 과정에서 발생하던 밝은 지표면에 대한 과도한 마스킹의 문제점을 개선시킨 것을 확인하였다.

• International Journal of Computer Science & Network Security
• /
• 제21권7호
• /
• pp.17-34
• /
• 2021

This study aimed to examine the Factor Structure of the teacher satisfaction scale (TSS) with distance education during the Covid-19 pandemic, as well as affirming the (Factorial Invariance) according to gender variable. It also aimed at identifying the degree of satisfaction according to some demographic variables of the sample. The study population consisted of all teachers in public education and faculty members in higher education in the Kingdom of Saudi Arabia. The (TSS) was applied to a random sample representing the study population consisting of (2399) respondents. The results of the study showed that the scale consists of five main factors, with a reliability value of (0.94). The scale also showed a high degree of construct validity through fit indices of the confirmatory factor analysis. The results have shown a gradual consistency of the measure's invariance that reaches the third level (Scalar-invariance) of the Measurement Invariance across the gender variable. The results also showed that the average response of the study sample on the scale reached (3.74) with a degree of satisfaction, as there are no statistically significant differences between the averages of the study sample responses with respect to the gender variable. While there were statistically significant differences in the averages with respect to the variable of the educational level in favor of the middle school and statistically significant differences in the averages attributed to the years of experience variable in favor of those whose experience is less than (5) years.

• 대한원격탐사학회지
• /
• 제27권2호
• /
• pp.191-201
• /
• 2011

본 연구에서는 다편파 레이더 산란계 자동 측정시스템 이용하여 콩 생육변화를 관측하고 레이더 시스템에서 얻어진 후방산란계수과 콩 생육인자들과의 관계분석을 통하여 콩 생육추정 가능성을 모색하고자 하였다. 2010년도 농촌진흥청 국립식량과학원 연구지역에 다편파 레이더 산란계 관측시스템 (L, C, X-밴드 안테나, 네트워크분석기, RF switch, 입사각 $40^{\circ}$)을 구축하고 콩 파종시기에서 수확기까지 10분단위로 콩 생육변화를 자동 측정하였다. 모든 안테나 밴드, 편파에서 콩 생육초기 (6월초~7월 중순)에는 VV-편파가 HH-, HV-편파보다 후방산란계수가 높게 나타났고, 그 이후 HH-편파와 다른 편파들 간의 cross-over 현상이 일어났는데 그 시기는 L-밴드가 7월 20일 (DOY 200), C-, X-밴드의 경우에는 7월 30일 (DOY 210)로 밴드에 따라 차이를 보였다. 모든 밴드 및 편파에서 9월 29일 (DOY 271)까지 후방산란계수가 증가하다가 그 이후 감소하였고 특히 종실비대기 (DOY 277, R6) 이후 감소폭이 크게 나타났는데 이 현상은 콩 생육인자 (초장, 엽면적지수, 건물중 등)변화와 일치하였다. 밴드에 따른 후방산란계수와 콩 생육인자들과의 관계를 분석한 결과 L-밴드 HH-편파에서 LAI ($r=0.93^{***}$), 초장 ($r=0.95^{***}$), 건물중 ($r=0.94^{***}$), 꼬투리중 ($r=0.92^{***}$)등 콩 생육인자들과의 상관계수가 가장 높게 나타났고 이에 비해 X-밴드 편파에서는 콩 생육인자들과의 상관계수가 상대적으로 낮게 나타났다. 후방산란계수 (L-밴드 HH-편파)를 이용하여 콩 생육인자 추정을 위한 회귀식을 작성하였다.

• 한국농림기상학회지
• /
• 제21권2호
• /
• pp.85-96
• /
• 2019

일사량은 자연 생태계와 농업 생태계에서 에너지 수지와 물 순환을 추정하는데 중요한 변수이다. 일별 일사량을 추정하기 위해 심층 신경망(DNN) 모델이 개발되었다. 일조시간 등의 변수보다 기상 관측소에서의 가용성이 더 높은 온도와 강수량이 심층 신경망 모델의 입력 자료로 사용되었다. five-fold crossvalidation 을 사용하여 심층 신경망을 훈련시키고 검증하였다. 국내 15 개의 기상 관측소에서 30 년 이상 장기간의 기상 자료가 수집되었다. Cross-validation을 통해 얻어진 심층 신경망 모델은 수원 지역 기상 관측소의 일별 일사량 추정치에 대해 비교적 작은 RMSE($3.75MJ\;m^{-2}\;d^{-1}$) 값을 가졌다. 심층 신경망 모델은 수원 지역 기상 관측소의 일사량의 변위의 약 68%를 설명했다. 1985 년과 1998 년의 일사량 관측값은 일조시간에 비해 상당히 낮은 값이 관측되었다. 이는 후속 연구에서 일사량 관측 데이터의 품질 평가가 필요할 것임을 시사했다. 해당 연도의 데이터를 분석에서 제외했을 때, 심층 신경망 모델의 추정값은 통계적 수치가 약간 높게 나타났다. 예를 들어, $R^2$ 와 RMSE 의 값은 각각 0.72 와 $3.55MJ\;m^{-2}\;d^{-1}$ 이었다. 심층 신경망 모델은 기온과 강수량을 통해 일사량을 추정하는데 유용하며, 이는 미래 기후 시나리오 자료에 대해서 활용할 수 있을 것이다. 따라서, 공간에 대한 제약이 완화된 심층 신경망 모델은 작물 모델의 입력 자료로 일사량이 필요한 작물 생산성에 대한 기후 변화 영향 평가에 유용하게 활용될 수 있을 것이다.