• 한국인터넷방송통신학회논문지
• /
• 제19권3호
• /
• pp.127-133
• /
• 2019

소프트웨어 결함 예측 연구들의 대부분은 라벨 데이터를 훈련 데이터로 사용하는 감독형 모델에 관한 연구들이다. 감독형 모델은 높은 예측 성능을 지니지만 대부분 개발 집단들은 충분한 라벨 데이터를 보유하고 있지 않다. 언라벨 데이터만 훈련에 사용하는 비감독형 모델은 모델 구축이 어렵고 성능이 떨어진다. 훈련 데이터로 라벨 데이터와 언라벨 데이터를 모두 사용하는 세미 감독형 모델은 이들의 문제점을 해결한다. Self-training은 세미 감독형 기법들 중 여러 가정과 제약조건들이 가장 적은 기법이다. 본 논문은 Self-training 알고리즘들을 이용해 여러 모델들을 구현하였으며, Accuracy와 AUC를 이용하여 그들을 평가한 결과 YATSI 모델이 가장 좋은 성능을 보였다.

• Journal of Information Processing Systems
• /
• 제17권3호
• /
• pp.453-461
• /
• 2021

Many deep learning approaches have been studied for image classification in computer vision. However, there are not enough data to generate accurate models in medical fields, and many datasets are not annotated. This study presents a new method that can use both unlabeled and labeled data. The proposed method is applied to classify cervix images into normal versus cancerous, and we demonstrate the results. First, we use a patch self-supervised learning for training the global context of the image using an unlabeled image dataset. Second, we generate a classifier model by using the transferred knowledge from self-supervised learning. We also apply attention learning to capture the local features of the image. The combined method provides better performance than state-of-the-art approaches in accuracy and sensitivity.

• 한국인터넷방송통신학회논문지
• /
• 제20권4호
• /
• pp.107-113
• /
• 2020

매우 많은 소프트웨어 결함 예측에 관한 연구들이 수행되어왔지만 대부분은 라벨 데이터를 훈련 데이터로 사용하는 감독형 모델들이었다. 언라벨 데이터만을 사용하는 비감독형 모델이나 언라벨 데이터와 매우 적은 라벨 데이터 정보를 함께 사용하는 세미감독형 모델에 관한 연구는 극소수에 불과하다. 본 논문은 Self-training 기법에 트리 알고리즘들을 사용하여 새로운 세미감독형 모델들을 제작하였다. 세미감독형 기법인 Self-training 모델에 트리 기법들을 사용하는 새로운 세미감독형 모델들을 제작하였다. 모델 평가 실험 결과 새롭게 제작한 트리 모델들이 기존 모델들보다 더 나은 성능을 보였으며, 특히 CollectiveWoods는 타 모델들에 비해 압도적으로 우월한 성능을 보였다. 또한 매우 적은 라벨 데이터 보유 상황에서도 매우 안정적인 성능을 보였다.

• Journal of Information Science Theory and Practice
• /
• 제1권1호
• /
• pp.10-26
• /
• 2013

Domain transfer is a widely recognized problem for machine learning algorithms because models built upon one data domain generally do not perform well in another data domain. This is especially a challenge for tasks such as opinion classification, which often has to deal with insufficient quantities of labeled data. This study investigates the feasibility of self-training in dealing with the domain transfer problem in opinion classification via leveraging labeled data in non-target data domain(s) and unlabeled data in the target-domain. Specifically, self-training is evaluated for effectiveness in sparse data situations and feasibility for domain adaptation in opinion classification. Three types of Web content are tested: edited news articles, semi-structured movie reviews, and the informal and unstructured content of the blogosphere. Findings of this study suggest that, when there are limited labeled data, self-training is a promising approach for opinion classification, although the contributions vary across data domains. Significant improvement was demonstrated for the most challenging data domain-the blogosphere-when a domain transfer-based self-training strategy was implemented.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제15권8호
• /
• pp.2993-3010
• /
• 2021

The job shop scheduling problem (JSSP) plays a critical role in smart manufacturing, an effective JSSP scheduler could save time cost and increase productivity. Conventional methods are very time-consumption and cannot deal with complicated JSSP instances as it uses one optimal algorithm to solve JSSP. This paper proposes an effective scheduler based on deep learning technology named self-supervised long-short term memory (SS-LSTM) to handle complex JSSP accurately. First, using the optimal method to generate sufficient training samples in small-scale JSSP. SS-LSTM is then applied to extract rich feature representations from generated training samples and decide the next action. In the proposed SS-LSTM, two channels are employed to reflect the full production statues. Specifically, the detailed-level channel records 18 detailed product information while the system-level channel reflects the type of whole system states identified by the k-means algorithm. Moreover, adopting a self-supervised mechanism with LSTM autoencoder to keep high feature extraction capacity simultaneously ensuring the reliable feature representative ability. The authors implemented, trained, and compared the proposed method with the other leading learning-based methods on some complicated JSSP instances. The experimental results have confirmed the effectiveness and priority of the proposed method for solving complex JSSP instances in terms of make-span.

• Journal of information and communication convergence engineering
• /
• 제22권2호
• /
• pp.165-171
• /
• 2024

In this study, we present a novel approach for enhancing chest X-ray image classification (normal, Covid-19, edema, mass nodules, and pneumothorax) by combining contrastive learning and machine learning algorithms. A vast amount of unlabeled data was leveraged to learn representations so that data efficiency is improved as a means of addressing the limited availability of labeled data in X-ray images. Our approach involves training classification algorithms using the extracted features from a linear fine-tuned Momentum Contrast (MoCo) model. The MoCo architecture with a Resnet34, Resnet50, or Resnet101 backbone is trained to learn features from unlabeled data. Instead of only fine-tuning the linear classifier layer on the MoCopretrained model, we propose training nonlinear classifiers as substitutes for softmax in deep networks. The empirical results show that while the linear fine-tuned ImageNet-pretrained models achieved the highest accuracy of only 82.9% and the linear fine-tuned MoCo-pretrained models an increased highest accuracy of 84.8%, our proposed method offered a significant improvement and achieved the highest accuracy of 87.9%.

• 대한방사선기술학회지:방사선기술과학
• /
• 제44권6호
• /
• pp.655-661
• /
• 2021

Positron emission tomography (PET) images is affected by acquisition time, short acquisition times results in low gamma counts leading to degradation of image quality by statistical noise. Noise2Void(N2V) is self supervised denoising model that is convolutional neural network (CNN) based deep learning. The purpose of this study is to evaluate denoising performance of N2V for PET image with a short acquisition time. The phantom was scanned as a list mode for 10 min using Biograph mCT40 of PET/CT (Siemens Healthcare, Erlangen, Germany). We compared PET images using NEMA image-quality phantom for standard acquisition time (10 min), short acquisition time (2min) and simulated PET image (S2 min). To evaluate performance of N2V, the peak signal to noise ratio (PSNR), normalized root mean square error (NRMSE), structural similarity index (SSIM) and radio-activity recovery coefficient (RC) were used. The PSNR, NRMSE and SSIM for 2 min and S2 min PET images compared to 10min PET image were 30.983, 33.936, 9.954, 7.609 and 0.916, 0.934 respectively. The RC for spheres with S2 min PET image also met European Association of Nuclear Medicine Research Ltd. (EARL) FDG PET accreditation program. We confirmed generated S2 min PET image from N2V deep learning showed improvement results compared to 2 min PET image and The PET images on visual analysis were also comparable between 10 min and S2 min PET images. In conclusion, noisy PET image by means of short acquisition time using N2V denoising network model can be improved image quality without underestimation of radioactivity.

• Korean Journal of Radiology
• /
• 제24권6호
• /
• pp.541-552
• /
• 2023

Objective: Detection of pneumoperitoneum using abdominal radiography, particularly in the supine position, is often challenging. This study aimed to develop and externally validate a deep learning model for the detection of pneumoperitoneum using supine and erect abdominal radiography. Materials and Methods: A model that can utilize "pneumoperitoneum" and "non-pneumoperitoneum" classes was developed through knowledge distillation. To train the proposed model with limited training data and weak labels, it was trained using a recently proposed semi-supervised learning method called distillation for self-supervised and self-train learning (DISTL), which leverages the Vision Transformer. The proposed model was first pre-trained with chest radiographs to utilize common knowledge between modalities, fine-tuned, and self-trained on labeled and unlabeled abdominal radiographs. The proposed model was trained using data from supine and erect abdominal radiographs. In total, 191212 chest radiographs (CheXpert data) were used for pre-training, and 5518 labeled and 16671 unlabeled abdominal radiographs were used for fine-tuning and self-supervised learning, respectively. The proposed model was internally validated on 389 abdominal radiographs and externally validated on 475 and 798 abdominal radiographs from the two institutions. We evaluated the performance in diagnosing pneumoperitoneum using the area under the receiver operating characteristic curve (AUC) and compared it with that of radiologists. Results: In the internal validation, the proposed model had an AUC, sensitivity, and specificity of 0.881, 85.4%, and 73.3% and 0.968, 91.1, and 95.0 for supine and erect positions, respectively. In the external validation at the two institutions, the AUCs were 0.835 and 0.852 for the supine position and 0.909 and 0.944 for the erect position. In the reader study, the readers' performances improved with the assistance of the proposed model. Conclusion: The proposed model trained with the DISTL method can accurately detect pneumoperitoneum on abdominal radiography in both the supine and erect positions.

• Asia pacific journal of information systems
• /
• 제29권4호
• /
• pp.789-816
• /
• 2019

Text classification is a challenging task, especially when dealing with a huge amount of text data. The performance of a classification model can be varied depending on what type of words contained in the document corpus and what type of features generated for classification. Aside from proposing a new modified version of the existing algorithm or creating a new algorithm, we attempt to modify the use of data. The classifier performance is usually affected by the quality of learning data as the classifier is built based on these training data. We assume that the data from different domains might have different characteristics of noise, which can be utilized in the process of learning the classifier. Therefore, we attempt to enhance the robustness of the classifier by injecting the heterogeneous data artificially into the learning process in order to improve the classification accuracy. Semi-supervised approach was applied for utilizing the heterogeneous data in the process of learning the document classifier. However, the performance of document classifier might be degraded by the unlabeled data. Therefore, we further proposed an algorithm to extract only the documents that contribute to the accuracy improvement of the classifier.

• 한국터널지하공간학회 논문집
• /
• 제21권3호
• /
• pp.419-432
• /
• 2019

대부분 딥러닝 모델의 학습은 입력값과 입력값에 따른 출력값이 포함된 레이블링 데이터(labeling data)를 학습하는 지도 학습(supervised learning)으로 진행된다. 레이블링 데이터는 인간이 직접 제작하므로 데이터의 정확도가 높다는 장점이 있지만 비용과 시간의 문제로 인해 데이터의 확보에 많은 노력이 소요된다. 그리고 지도 학습의 목표는 정탐지 데이터(true positive data)의 인식 성능 향상에 초점이 맞추어져 있으며, 오탐지 데이터(false positive data)의 발생에 대한 대처는 미흡한 실정이다. 본 논문은 터널 관제센터에 투입된 딥러닝 모델 기반 영상유고 시스템의 모니터링을 통해 정탐지와 레이블링 데이터의 학습으로 예측하기 힘든 오탐지의 발생을 확인하였다. 오탐지의 유형은 작업차량의 경광등, 터널 입구부에서 반사되는 햇빛, 차선과 차량의 일부에서 발생하는 길쭉한 검은 음영 등이 화재와 보행자로 오탐지되고 있었다. 이러한 문제를 해결하기 위해 현장에서 발생한 오탐지 데이터와 레이블링 데이터를 동시에 학습하여 딥러닝 모델을 개발하였으며, 그 결과 기존 레이블링 데이터만 학습한 모델과 비교하면 레이블링 데이터에 대한 재추론 성능이 향상됨을 알 수 있었다. 그리고 오탐지 데이터에 대한 재추론을 한 결과 오탐지 데이터를 많이 포함하여 학습한 모델일 경우 보행자의 오탐지 개수가 훨씬 줄었으며, 오탐지 데이터의 학습을 통해 딥러닝 모델의 현장 적용성을 향상시킬 수 있었다.