하수처리장의 안정성과 효율성의 향상을 위해 스마트 기술 도입이 요구되고 있으나, 운영 데이터베이스 구축에 있어 계측의 신뢰성과 연속성 확보에 어려움이 있다. 활성슬러지 모델은 하수처리장의 디지털트윈으로 활용되며, 유입수 성상이 동일하더라도 다양한 운전 조건에 대한 데이터를 생산할 수 있다. 본 연구에서는 실측 데이터와 시뮬레이터 기반 합성 데이터를 통합하여 하수처리장 질소 농도 예측 머신러닝 모델을 구축하였다. A2O 공정의 호기조를 대상으로 기체상 N2O 및 액상 NH4 + 농도를 측정하였으며, 내부반송량, 외부반송량 등 운전인자를 포함한 운영데이터베이스를 구축하고 분석하였다. 확보한 실측 데이터를 기반으로 운영 특성을 분석하고, Sumo4N 모델을 활용하여 다양한 운전 조건에서의 합성 데이터를 생성하였다. 이후 두 데이터를 통합하여 데이터 증강을 수행함으로써, 실측 데이터의 양적 한계를 보완하였다. 모델 학습을 위한 입력 변수로는 외부⋅내부 반송량, 폭기량, 온도, 유입 질소 부하, pH를 선정하였으며 호기조의 N2O, NH4 +과 방류수 TN 농도를 예측하기 위한 머신러닝 모델을 개발하였다. 모델 학습에는 Lasso Regression, Random Forest, k-NN, SVR 알고리즘을 적용하여 성능을 평가하였다. 그 결과 SVR 알고리즘이 모든 질소 성분 예측에서 가장 우수한 성능을 보였으며, 개발된 모델 모두 R² ≥ 0.75의 높은 예측 성능을 나타내었다. 이는 시뮬레이터 기반 데이터 증강을 통해 기체상 및 액상 질소의 통합 제어를 위한 머신러닝 모델 구축의 가능성을 시사한다.

본 연구에서는 작물 이미지 데이터셋에 적합한 데이터 증강 기법으로 자연광 기반 증강을 제안하였다. NLA는 밝기, 색온 도, 대비 변화를 반영하여 실제 환경에서 발생 가능한 광조건 을 학습 데이터에 적용함으로써, 기존 YOLO 기본 증강 대비 더욱 현실적인 학습 데이터를 생성할 수 있도록 설계되었다. 참외와 딸기 데이터셋을 대상으로 한 실험 결과, NLA는 Baseline 대비 전반적인 성능 향상을 보였다. 또한 Baseline에 서 학습 후반부에 나타난 성능 저하나 불안정성이 NLA에서 는 완화되어, 제안 기법이 과적합 방지와 학습 안정성 유지에 도 효과적임을 입증하였다. PatchSwap은 Mosaic의 한계를 보완하기 위한 보조적 시도로 실험에 포함되었으나, 종합적 인 성능 향상에는 뚜렷한 기여를 하지 못하였다. 본 연구 결과는 작물 이미지에서 실제 광조건을 반영한 데 이터 증강의 필요성과 유효성을 보여주었으며, 이는 향후 다 양한 작물과 촬영 환경으로 확장 가능한 가능성을 제시한다. 앞으로는 NLA의 적용 범위를 확대하고, 다른 딥러닝 모델 및 추가적인 농업 데이터셋에 대한 성능 검증을 통해 본 기법의 범용성과 실효성을 더욱 강화할 필요가 있다.

전 세계 식량 안보는 기후 변화와 인구 증가로 인해 점점 더 위협받고 있으며, 이를 해결하기 위해서는 유전체학, 표현형 학, 인공지능을 통합한 첨단 육종 전략이 필요하다. 본 연구는 유전자형 데이터 증강과 반지도 학습을 활용하여 토마토 육종 에서의 표현형 예측 정확도를 향상시키는 것을 목표로 한다. 총 192종의 토마토 계통을 온실 환경에서 재배하며, 과중, 높 이, 너비, 경도, 당도 등 5가지 주요 형질에 대한 유전자형, 표 현형, 환경 데이터를 수집한다. 제안된 1차원 합성곱신경망 기반의 유전자형 데이터 증강 프레임워크는 원본 데이터셋을 확장하고, 라벨이 안된 데이터를 효과적으로 활용하기 위한 수도 라벨링 전략을 도입한다. 또한, 온도, 습도 등 환경 변수 는 생육 기간 동안의 통계적 특징값을 추출하여 모델 입력에 통합함으로써 재배 조건을 보다 현실적으로 반영하였다. 표 현형 예측은 트리 기반 및 딥러닝 아키텍처를 포함한 다양한 모델을 통해 수행되었으며, 서로 다른 네트워크 구조에 따른 성능을 비교 및 평가한다. 실험 결과, 유전자형 데이터 증강은 전반적으로 예측 성능을 향상시켰으며, 특히 LightGBM과 CatBoost와 같은 트리 기반 모델에서 가장 큰 개선 효과를 보 였다. 또한 최신 딥러닝 모델과의 비교 실험을 통해 제안된 접 근법의 강건성을 확인한다. 이러한 결과는 제안된 방법이 데 이터가 제한된 육종 환경에서도 실질적인 성능 향상을 달성할 수 있는 효과적인 전략임을 보여주며, 향후 멀티오믹스 및 환 경 데이터와의 통합을 통해 확장 가능한 디지털 육종 프레임 워크로 발전할 가능성을 제시한다.

In case of gingival recession and alveolar bone defects due to tooth loss for a long period of time in a single tooth in the maxillary anterior region, it is not easy to obtain aesthetic results with a single implant prosthesis. For aesthetic restoration, it is important to preserve hard and soft tissues through alveolar bone augmentation as well as restore harmony with adjacent teeth and soft tissues by placing the implant in an ideal location. In this case, an implant was placed using guided bone regeneration and a connective tissue graft simultaneously with immediate implantation after extraction from the maxillary anterior region where only residual root was left for a long period of time.

In this paper, a GAN-based data augmentation method is proposed for topology optimization. In machine learning techniques, a total amount of dataset determines the accuracy and robustness of the trained neural network architectures, especially, supervised learning networks. Because the insufficient data tends to lead to overfitting or underfitting of the architectures, a data augmentation method is need to increase the amount of data for reducing overfitting when training a machine learning model. In this study, the Ganerative Adversarial Network (GAN) is used to augment the topology optimization dataset. The produced dataset has been compared with the original dataset.

The manner in which education will be delivered in the 21st-century has often been debated. Various literature has agreed that an interactive teaching and learning method is required in parallel with the emergence and development of cyber technology. The conventional method of teaching should be reconstituted to emphasize aspects associated with innovation and creativity in attracting the attention of students in learning. Despite the current Malaysian education emphasize the learning features that include 1) Creative thinking, 2) Critical thinking, 3) Collaboration, 4) Character and 5) Communication. However, 21st-century approach requires exposure, skills, and creativity to be implemented by the Malaysian educators. Therefore, the aim of this study is to propose a new maritime interactive teaching model towards a Sustainable Development Goals (SGDs) and industrial revolution 4.0. Three (3) secondary schools around Terengganu in Malaysia were chosen to participate in a pilot case study. The results of the study found that more than 90% of students now understand more about the maritime industry based on their acquired knowledge and education in this area. While, more than 70% of students described that this method of teaching is appealing. Maritime education innovative learning through an interactive learning model was successfully achieved based on the findings of this study, called the ‘Mariner’s Fantasy’. Additionally, through the inspirations of IR 4.0 and the Malaysia Education Development Plan, 2013-2025, the study has demonstrated the usefulness of the Maritime Education Innovative Learning (MEIL) program through an interactive learning method, in enhancing the delivery of maritime education by adopting an effective teaching-based approach.

Augmentative biological control, which refers to frequent releases of mass produced natural enemies, is well developedin the greenhouse cropping systems and many successful cases have been reported worldwide. The most robust biocontrolsystems are based on the principle of established populations of natural enemies before invasions of pests occur. However,the natural enemies that are commercially available do not establish well in all crops, which can be caused by a lackof food sources, a lack of oviposition sites or shelter, or because of an unsuitable microclimate. These problems can besolved by selecting new species of natural enemies that are better adapted to the crop and/or by providing the resourcesthat natural enemies need to survive and reproduce in the crop. Here we present some results of studies that may contributeto this “standing army approach” in biocontrol, based on the use of banker plants, nectar plants, rearing sachets, foodsprays and alternative prey applications.in the greenhouse cropping systems and many successful cases have been reported worldwide. The most robust biocontrolsystems are based on the principle of established populations of natural enemies before invasions of pests occur. However,the natural enemies that are commercially available do not establish well in all crops, which can be caused by a lackof food sources, a lack of oviposition sites or shelter, or because of an unsuitable microclimate. These problems can besolved by selecting new species of natural enemies that are better adapted to the crop and/or by providing the resourcesthat natural enemies need to survive and reproduce in the crop. Here we present some results of studies that may contributeto this “standing army approach” in biocontrol, based on the use of banker plants, nectar plants, rearing sachets, foodsprays and alternative prey applications.

Computational results with pseudoplastic fluid flows for fully developed non-Newtonian laminar flows have been obtained. Those consist of the product of friction factor and Modified Reynolds number and Nusselt numbers with respect to the shear rate parameter in an annular pipe. The numerical results of the product of friction factor and Reynolds numbers and the Nusselt numbers for both Newtonian region and the power law region were compared with previously published asymptotic results, respectively. In the present calculations, the product of friction factor and Newtonian Reynolds numbers for pseudoplastic fluid at power law region in annular pipe is 180% less than that for Newtonian fluid. For power law fluids with different power law flow indices, the difference of the product of friction factor and power law Reynolds number between previous and the present results at the power law region is within 0.20%. The solutions also show the effect of the shear rate parameter on the Nusselt number and about 11% increase of Nusselt number at the power region.

The present study was carried out to establish an animal model, displaying long-term learning and memory dysfunction, since single intracerebroventricular (icv) injection of amyloid β peptide (Aβ) causes a short-term memory impairment. Male ICR mice were fed a high-cholesterol diet (HCD) containing 3% cholesterol, 1% corn oil and 0.5% cholic acid, and 1 week later, icv injected with Aβ_1-42 (5 μg/head). Learning/ memory function was assessed via passive avoidance performances 1 day and 2, 4, and 6 weeks after Aβ_1-42 injection, in addition to blood biochemical analyses for lipid profiles and hepatic function. Total cholesterol, lowdensity lipoproteins and hepatic dysfunction parameters markedly increased, while high-density lipoproteins were reduced following HCD feeding. Whereas single injection of Aβ induced temporary memory loss 1 day after administration, exhibiting full recovery after 2 weeks, Aβ treatment in combination with HCD feeding lasted the learning/memory impairment up to 6 weeks. Therefore, it is suggested that hypercholesterolemia augments Aβ-induced memory loss, and that Aβ injection plus HCD feeding could be a long-term memorydeficit model suitable for long-term treatment with drugs or stem cells.

동충하초로 불리우는 곤충병원진균(Paecilomycesjaponicus)이 의약적으로 상품화되어 사용되고 있으며,누에(Bombyxmori)가 이 진균의 최적 기주로 선발되어 자실테 생산에 이용되고 있다. 현재 이 균주의 처리는 갓 탈피한 누에 5령 유충에 접종하고 고온(30\"C),다습(약 90%상대 습도)및 24시간 절식 조건에서 스트레스에 의한 면역 저하를 유도하여 균주 접종율을 높히는 방식을 취하고 있다. 본 연구는 면역반응 중개에 중요한 eicosanoid반응을 억제시키는 dexamethas-one(DEX)을 이용하여 물리적 스트레스 환경의 조성 없이도 누에에 면역 저하를 유도시키려는 목적으로 수행되었다. 누에 령 유충에 주입된 DEX(1007g)는 병원진균의 혈구치사 능력을 뚜렷 이 증가시켰다. 또 DEX(1007g)는 작은혹형성이나 피막형성에서 나타나는 혈구응집 반응이나, phenoloxidase활성으로 측정된 누에의 세포성 면역 반응을 뚜렷이 저하시켰다 효과적 병원진균 의 충체 처리를 위해 곤충체의 부착 능력을 제고시켜 접종율을 높히는 것으로 본 연구에서 판명된 Triton-X(0.05%)를 모든 충체 처리 용액에 이용되었다. DEX(100)단독처리가 기존의 물리 적 스트레스 환경 처리를 통한 방법과 유사한 수준으로 병원진균의 접종율을 나타냈다. 본 연구는 DEX가 동충하초 접종율을 제고시킬 수 있음을 시사했고, 누에는 이러한 진균 병원체에 대해 서 eicosanoid를 이용하여 세포성 면역을 발현하는 것으로 제시하고 있다.고 있다.

This paper is a study on data augmentation for small dataset by using deep learning. In case of training a deep learning model for recognition and classification of non-mainstream objects, there is a limit to obtaining a large amount of training data. Therefore, this paper proposes a data augmentation method using perspective transform and image synthesis. In addition, it is necessary to save the object area for all training data to detect the object area. Thus, we devised a way to augment the data and save object regions at the same time. To verify the performance of the augmented data using the proposed method, an experiment was conducted to compare classification accuracy with the augmented data by the traditional method, and transfer learning was used in model learning. As experimental results, the model trained using the proposed method showed higher accuracy than the model trained using the traditional method.

In the ground environment, mobile robot research uses sensors such as GPS and optical cameras to localize surrounding landmarks and to estimate the position of the robot. However, an underwater environment restricts the use of sensors such as optical cameras and GPS. Also, unlike the ground environment, it is difficult to make a continuous observation of landmarks for location estimation. So, in underwater research, artificial markers are installed to generate a strong and lasting landmark. When artificial markers are acquired with an underwater sonar sensor, different types of noise are caused in the underwater sonar image. This noise is one of the factors that reduces object detection performance. This paper aims to improve object detection performance through distortion and rotation augmentation of training data. Object detection is detected using a Faster R-CNN.

India is the top recipient of workers’ remittance flows; recent data indicate that the Remittances/GDP ratio has increased from 2.7% in 2000 to 3.36% in 2015. We apply a consumption behavior model, based on the “permanent income hypothesis”, to estimate the consumption augmentation and the stability impact for the period of 1989-2014. The independent variables are: (i) real per capita income (exclusive of remittances) is the measure of “permanent income”, (ii) remittances is the measure of “transitory income”, and (iii) real interest rate as the indicator of consumers’ ability for intertemporal consumption. The economic ramifications are important since current global risk factors could decrease flows in the future. The results indicate the significance of all three variables; there are: (i) evidence of significant consumption augmentation, (ii) consumption responds higher to remittances than to real income, the remittance elasticity is 0.571 and the income elasticity is 0.31, and (iii) evidence of pro-cyclical effect. The VAR model indicates some linkages and causality in the series that result in small response to the shocks. Policies to increase or stabilize remittance flows and to leverage remittances for economic development are important.

Recently severe drought caused the water shortage around the western parts of Chungcheongnamdo province, South Korea. A Diversion tunnel from the Geum river to the Boryong dam, which is the water supply dam for these areas has been proposed to solve this problem. This study examined hydraulic impacts on the Geum river associated with the diversion plan assuming the severe drought condition of 2015 would persist for the simulation period of 2016. The hydraulic simulation model was verified using hydrologic and hydraulic data including hourly discharges of the Geum river and its 8 tributaries, fluctuation of tidal level at the mouth of the river, withdrawals and return flows and operation records of the Geum river barrage since Feb. 1, 2015 through May 31, 2015. For the upstream boundary condition of the Geum river predicted inflow series using the nonlinear regression equation for 2015 discharge data was used. In order to estimate the effects of uncertainty in inflow prediction to the results total four inflow series consisting of upper limit flow, expected flow, lower limit flow and instream flow were used to examine hydraulic impacts of the diversion plan. The simulation showed that in cases of upper limit and expected flows there would be no problem in taking water from the Geum river mouth with a minimum water surface level of EL(+) 1.44 m. Meanwhile, the simulation also showed that in cases of lower limit flow and instream flow there would be some problems not only in taking water for water supply from the mouth of the Geum river but also operating the diversion facility itself with minimum water surface levels of EL(+) 0.94, 0.72, 0.43, and 0.14 m for the lower limit flow without/with diversion and the instream flow without/with diversion, respectively.

광역보정시스템은 GPS와 같은 위성항법을 이용하는 사용자의 정확성, 무결성을 개선시키기 위하여 고안된 시스템이다. 본 논문에서 는 개발된 의사위성 기반의 광역보정시스템의 전체 구조에 대하여 설명하고, 후처리 기반으로 상용 수신기에 대하여 성능 테스트를 수행하는 실험 방법 및 결과에 대하여 설명한다. 보정정보 생성을 위하여 총 5개의 NDGPS 기준국에서 수집되는 데이터가 활용되었으며 이를 광역기준 국, 중앙처리국 소프트웨어에서 처리하였다. 생성된 보정정보는 SP3, IONEX 데이터와 비교하여 정확도를 테스트하였다. 상용 수신기 실험에서 는 사용자의 RF 신호를 수집, 보정정보를 생성하였으며, 이후에 RF신호와 보정정보가 실린 의사위성 신호를 동시에 방송하여 테스트를 수행하 였다. 테스트는 3대의 상용수신기를 활용하여 수행되었으며 MSAS, GPS 단독 측위 수신기와 비교하여 성능을 검증하였다. 각 수신기의 위치 해 출력 결과로부터 위치오차를 비교하였으며 보정정보를 적용함으로써 향상된 위치해가 출력됨을 확인하였다.