• KSBB Journal
• /
• 제27권5호
• /
• pp.295-300
• /
• 2012

Matured dendritic cells (DCs) begin migration with their release from the bone marrow (BM) into the blood and subsequent traffic into peripheral lymphoid and non-lymphoid tissues. Throughout this long movement, migrating DCs must apply specialized skills to reach their target destination. Non-invasive in vivo cell-tracking techniques are necessary to advance immune cell-based therapies. In this study, we used a DiD cell-tracking solution for in vivo dendritic cell tracking in naive mice. We tracked DiD (non-invasive fluorescence dye)-labeled mature dendritic cells using the Near Infrared (NIR) imaging system in normal mice. We examined the immunophenotype of DiD-labeled cells compared with non-labelled mature DCs, and obtained time-serial images of NIR-DC trafficking after mouse footpad injection. In conclusion, we confirmed that DiD-labeled DCs migrated into the popliteal lymph node 24 h after the footpad injection. Here, these data suggested that the cell tracking system with the stable fluorescence dye DiD was useful as a cell tracking tool to advance dendritic cell-based immunotherapy.

• 한국응용약물학회:학술대회논문집
• /
• 한국응용약물학회 1995년도 춘계학술대회
• /
• pp.80-80
• /
• 1995

T4 Endonuclease V (Mw 16,000) acts as a repair enzyme for UV induced pyrimidine dimers in DNA. Many researchers have studied the biochemical characteristics of the enzyme. However the precise action mechanism of T4 endo V has not fully elucidated yet. In our laboratory NMR spectroscopy technique is being used for the structural study of T4 endo V. Because of its low temperature stability and high content of ${\alpha}$-helix, the conventional $^1$H NMR technique was inapplicable. Therefore we utilized stable isotope labeling technique and so far prepared about 10 amino acid specific labeled proteins. The HSQC spectra of amino acid specific labeled proteins will help us to interpret the triple resonance 3D, 4D data which are under processing, We also studied the behaviors of specific amino acid residues whose roles might be critical. When the enzyme labeled by $\^$15/N-Thr was mixed with the substrate oligonucleotide (semispecific -TT- sequence), one crosspeak in its HSQC spectrum was completely desappeared, which means that one of seven Thr residues is in the binding site of the enzyme with DNA, This result is well consistent with previous report that implicated the Thr 2 residue in the activity of the enzyme. Similar studies were carried on the behaviors of Arg and Tyr residues.

• Journal of the Korean Data and Information Science Society
• /
• 제25권2호
• /
• pp.447-454
• /
• 2014

In many practical machine learning and data mining applications, unlabeled training examples are readily available but labeled ones are fairly expensive to obtain. Therefore semi-supervised learning algorithms have attracted much attentions. However, previous research mainly focuses on classication problems. In this paper, a semi-supervised regression method based on support vector regression (SVR) formulation that is proposed. The estimator is easily obtained via the dual formulation of the optimization problem. The experimental results with simulated and real data suggest superior performance of the our proposed method compared with standard SVR.

• International Journal of Oral Biology
• /
• 제41권2호
• /
• pp.97-103
• /
• 2016

Mammals have 3 pairs of major salivary glands i.e., the parotid, submandibular, and sublingual glands. Saliva secretion of these glands is modulated by taste perception. Salivary glands are composed mainly of acinar and ductal cells. Primary saliva is secreted by acinar cells and modified during ductal flow. Recently, of the murine 35 bitter taste receptors, Tas2r108 was expressed at highest levels in the submandibular gland by qPCR. Further, Tas2r108-transfected cells respond to a range of bitter compounds, such as denatonium, quinine, colchicine, diphenidol, caffeine and dapson. The objective of the present study was to characterize the expression of Tas2r108 mRNA in acinar and/or ductal cells of the submandibular gland using in situ hybridization (ISH). Male 42-60 days old DBA2 mice were used in the study. Messenger RNAs were extracted from the submandibular gland for generating digoxigenin (DIG) labeled-cRNA probes. These probes were transcribed in anti-sense and sense orientation using T7 RNA polymerase. Dot blot hybridization was performed using DIG labeled-cRNA probes, in order to estimate integrity and optimal diluting concentration of these probes. Subsequently, ISH was performed on murine submandibular gland to detect Tas2r108 mRNA. Dot blot hybridization data demonstrated that Tas2r108 DIG labeled-cRNA anti-sense probes specifically detected Tas2r108 cDNA. ISH results showed that the anti-sense probes labeled acinar and ductal cells in the submandibular gland, whereas no staining was visible in sense controls. Interestingly, the Tas2r108 expression levels were higher in acinar than ductal cells. These results suggested that Tas2r108 might be more associated with primary saliva secretion than with ductal modification of saliva composition.

• 수산해양교육연구
• /
• 제22권4호
• /
• pp.611-620
• /
• 2010

The purpose of this study was to analysis of the image of elementary school teachers on special class. The Q-methodology was used as a research method which is useful for analyzing subjectivity like as human value, attitude, belief and image. P-sample as subjects of this study was collected from 40 elementary school teacher in Busan. A data was analyzed using QUANL PC program. As a result of research, the image of elementary school teachers on special class was classified as 4 type. The 1 type was labeled the practical proponent type. The 1 type teachers highly estimate the educational importance of the special class and thought it as a place that educational possibility of children with disabilities can come true. The 2 type was labeled the passive bystander type. They are quite indifferent to the special class, and are skeptical about the educational achievement of children with disabilities. The 3 type was labeled the ideational positive type. They admit the importance of the special class, but thought that the image about the special class highly depends on the ability of the teachers who take responsibility for it. The 4 type was labeled the realistic opponent. They criticize the special class for not being the place that inclusion education for children with disabilities comes true and they are dissatisfied with the fact that ordered-curriculum for children with disabilities isn't applied at all.

• 스마트미디어저널
• /
• 제11권4호
• /
• pp.30-37
• /
• 2022

Disease threatens plant growth and recognizing the type of disease is essential to making a remedy. In recent years, deep learning has witnessed a significant improvement for this task, however, a large volume of labeled images is one of the requirements to get decent performance. But annotated images are difficult and expensive to obtain in the agricultural field. Therefore, designing an efficient and effective strategy is one of the challenges in this area with few labeled data. Transfer learning, assuming taking knowledge from a source domain to a target domain, is borrowed to address this issue and observed comparable results. However, current transfer learning strategies can be regarded as a supervised method as it hypothesizes that there are many labeled images in a source domain. In contrast, unsupervised transfer learning, using only images in a source domain, gives more convenience as collecting images is much easier than annotating. In this paper, we leverage unsupervised transfer learning to perform plant disease recognition, by which we achieve a better performance than supervised transfer learning in many cases. Besides, a vision transformer with a bigger model capacity than convolution is utilized to have a better-pretrained feature space. With the vision transformer-based unsupervised transfer learning, we achieve better results than current works in two datasets. Especially, we obtain 97.3% accuracy with only 30 training images for each class in the Plant Village dataset. We hope that our work can encourage the community to pay attention to vision transformer-based unsupervised transfer learning in the agricultural field when with few labeled images.

• Smart Structures and Systems
• /
• 제29권6호
• /
• pp.757-766
• /
• 2022

To train deep learning algorithms, a sufficient number of data are required. However, in most engineering systems, the acquisition of fault data is difficult or sometimes not feasible, while normal data are secured. The dearth of data is one of the major challenges to developing deep learning models, and fault diagnosis in particular cannot be made in the absence of fault data. With this context, this paper proposes an anomaly detection methodology for rotating machines using only normal data with self-labeling. Since only normal data are used for anomaly detection, a self-labeling method is used to generate a new labeled dataset. The overall procedure includes the following three steps: (1) transformation of normal data to self-labeled data based on a pretext task, (2) training the convolutional neural networks (CNN), and (3) anomaly detection using defined anomaly score based on the softmax output of the trained CNN. The softmax value of the abnormal sample shows different behavior from the normal softmax values. To verify the proposed method, four case studies were conducted, on the Case Western Reserve University (CWRU) bearing dataset, IEEE PHM 2012 data challenge dataset, PHMAP 2021 data challenge dataset, and laboratory bearing testbed; and the results were compared to those of existing machine learning and deep learning methods. The results showed that the proposed algorithm could detect faults in the bearing testbed and compressor with over 99.7% accuracy. In particular, it was possible to detect not only bearing faults but also structural faults such as unbalance and belt looseness with very high accuracy. Compared with the existing GAN, the autoencoder-based anomaly detection algorithm, the proposed method showed high anomaly detection performance.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제15권7호
• /
• pp.2304-2320
• /
• 2021

Intelligently detecting anomalies in health sensor data streams (e.g., Electrocardiogram, ECG) can improve the development of E-health industry. The physiological signals of patients are collected through sensors. Timely diagnosis and treatment save medical resources, promote physical health, and reduce complications. However, it is difficult to automatically classify the ECG data, as the features of ECGs are difficult to extract. And the volume of labeled ECG data is limited, which affects the classification performance. In this paper, we propose a Generative Adversarial Network (GAN)-based deep learning framework (called CAB) for heart arrhythmia classification. CAB focuses on improving the detection accuracy based on a small number of labeled samples. It is trained based on the class-imbalance ECG data. Augmenting ECG data by a GAN model eliminates the impact of data scarcity. After data augmentation, CAB classifies the ECG data by using a Bidirectional Long Short Term Memory Recurrent Neural Network (Bi-LSTM). Experiment results show a better performance of CAB compared with state-of-the-art methods. The overall classification accuracy of CAB is 99.71%. The F1-scores of classifying Normal beats (N), Supraventricular ectopic beats (S), Ventricular ectopic beats (V), Fusion beats (F) and Unclassifiable beats (Q) heartbeats are 99.86%, 97.66%, 99.05%, 98.57% and 99.88%, respectively. Unclassifiable beats (Q) heartbeats are 99.86%, 97.66%, 99.05%, 98.57% and 99.88%, respectively.

• 응용통계연구
• /
• 제34권2호
• /
• pp.167-175
• /
• 2021

준지도학습(semi-supervised learning)은 목표값이 있는 데이터와 없는 데이터를 모두 이용하는 학습방법이다. 준지도학습에서 최근에 많은 관심을 받는 일치성규칙(consistency regularization)과 데이터 증대를 이용한 준지도학습(unsupervised data augmentation; UDA)은 목표값이 없는 데이터를 증대하여 학습에 이용한다. 그리고 성능 향상을 위해 훈련신호강화(training signal annealing; TSA)와 신뢰기반 마스킹(confidence based masking)을 이용한다. 본 연구에서는 UDA에서 사용하는 KL-정보량(Kullback-Leibler divergence)과 TSA 대신 JS-정보량(Jensen-Shanon divergene)과 역-TSA를 사용하고 신뢰기반 마스킹을 제거하는 방법을 제안한다. 실험을 통해 제안된 방법의 성능이 더 우수함을 보였다.

• 한국지능시스템학회논문지
• /
• 제24권5호
• /
• pp.482-488
• /
• 2014

대표적인 소셜 네트워크 서비스(SNS)인 트위터의 내용을 분석하여 자동으로 트윗에 나타난 사용자의 감성을 분석하고자 한다. 기계학습 기법을 사용해서 감성 분석 모델을 생성하기 위해서는 각각의 트윗에 긍정 또는 부정을 나타내는 감성 레이블이 필요하다. 그러나 사람이 모든 트윗에 감성 레이블을 붙이는 것은 비용이 많이 소요되고, 실질적으로 불가능하다. 그래서 본 연구에서는 "감성 레이블이 있는 데이터"와 함께 "감성 레이블이 없는 데이터"도 활용하기 위해서 반감독 학습기법인 self-training 알고리즘을 적용하여 감성분석 모델을 생성한다. Self-training 알고리즘은 "레이블이 있는 데이터"의 레이블이 있는 데이터를 활용하여 "레이블이 없는 데이터"의 레이블을 확정하여 "레이블이 있는 데이터"를 확장하는 방식으로, 분류모델을 점진적으로 개선시키는 방식이다. 그러나 데이터의 레이블이 한번 확정되면 향후 학습에서 계속 사용되므로, 초기의 오류가 계속적으로 학습에 영향을 미치게 된다. 그러므로 조금 더 신중하게 "레이블이 없는 데이터"의 레이블을 결정할 필요가 있다. 본 논문에서는 self-training 알고리즘을 이용하여 보다 높은 정확도의 감성 분석 모델을 생성하기 위하여, self-training 중 "감성 레이블이 없는 데이터"의 레이블을 결정하여 "감성 레이블이 있는 데이터"로 확장하기 위한 3가지 정책을 제시하고, 각각의 성능을 비교 분석한다. 첫 번째 정책은 임계치를 고려하는 것이다. 분류 경계로부터 일정거리 이상 떨어져 있는 데이터를 선택하고자 하는 것이다. 두 번째 정책은 같은 개수의 긍/부정 데이터를 추가하는 것이다. 한쪽 감성에 해당하는 데이터에만 국한된 학습을 하는 것을 방지하기 위한 것이다. 세 번째 정책은 최대 개수를 고려하는 것이다. 한 번에 많은 양의 데이터가 "감성 레이블이 있는 데이터"에 추가되는 것을 방지하고 상위 몇%만 선택하기 위해서, 선택되는 데이터의 개수의 상한선을 정한 것이다. 실험은 긍정과 부정으로 분류되어 있는 트위터 데이터 셋인 Stanford data set에 적용하여 실험하였다. 그 결과 학습된 모델은 "감성 레이블이 있는 데이터" 만을 가지고 모델을 생성한 것보다 감성분석의 성능을 향상 시킬 수 있었고 3가지 정책을 적용한 방법의 효과를 입증하였다.