• Journal of KIISE
• /
• v.41 no.12
• /
• pp.1110-1116
• /
• 2014

Nowadays, social networking sites (SNSs), such as Twitter, LinkedIn, and Tumblr, are coming into the forefront, due to the growth in the number of users. While users voluntarily provide their information in SNSs, privacy leakages resulting from the use of SNSs is becoming a problem owing to the evolution of large data processing techniques and the raising awareness of privacy. In order to solve this problem, the studies on protecting privacy on SNSs, based on graph and machine learning, have been conducted. However, examples of privacy leakages resulting from the advent of a new SNS are consistently being uncovered. In this paper, we propose a technique enabling a user to detect privacy leakages beforehand in the case where the service provider or third-party application developer threatens the SNS user's privacy maliciously.

• The Journal of Korean Institute of Electromagnetic Engineering and Science
• /
• v.18 no.10
• /
• pp.1157-1162
• /
• 2007

In this paper, E-TDLNN(Expanded-Tapped Delay Line Neural Network) method is suggested to make the model of power amplifier effectively. This method is the one for making the model of power amplifier through the study in neural network to the target value, the measured output spectrum of power amplifier, after adding the external value factor, gate bias, as an invariant input to the TDLNN method which suggested the memory effect of power amplifier effectively. To prove the validity of suggested method, the data at 2 points, 3.45 V and 3.50 V of gate bias range $3.4{\sim}3.6V$ with the 0.01 V step change, are studied and the predicted results at the gate bias 3.40 V, 3.48 V, 3.53 V and 3.60 V shows good coincidence with the measured values.

• Journal of the Institute of Electronics Engineers of Korea CI
• /
• v.41 no.3
• /
• pp.55-65
• /
• 2004

When a neural network is used to solve a given problem it is necessary to match the complexity of the network to that of the problem because the complexity of the network significantly affects its learning capability and generalization performance. Thus, it is desirable to have an algorithm that can find appropriate network structures in a self-organizing way. This paper proposes algorithms which automatically organize feed forward multi-layer neural networks with sigmoid hidden neurons for given problems. Using both constructive procedures and pruning procedures, the proposed algorithms try to find the near optimal network, which is compact and shows good generalization performance. The performances of the proposed algorithms are tested on four function regression problems. The results demonstrate that our algorithms successfully generate near-optimal networks in comparison with the previous method and the neural networks of fixed topology.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.24 no.1
• /
• pp.8-14
• /
• 2020

The human impact of particulate matter are revealed and demand for improved forecast accuracy is increasing. Recently, efforts is made to improve the accuracy of PM10 predictions by using machine learning, but prediction performance is decreasing due to the particulate matter data with a large rate of low concentration occurrence. In this paper, separation prediction model by concentration is proposed to improve the accuracy of PM10 particulate matter forecast. The low and high concentration prediction model was designed using the weather and air pollution factors in Cheonan, and the performance comparison with the prediction models was performed. As a result of experiments with RMSE, MAPE, correlation coefficient, and AQI accuracy, it was confirmed that the predictive performance was improved, and that 20.62% of the AQI high-concentration prediction performance was improved.

• The Journal of the Acoustical Society of Korea
• /
• v.42 no.4
• /
• pp.345-356
• /
• 2023

Target signal of passive sonar shows narrow band harmonic characteristic with a variation in intensity within a few seconds and long term frequency variation due to the Lloyd's mirror effect. We propose a signal classification algorithm based on Gated Recurrent Unit (GRU) that learns local and global time series features. The algorithm proposed implements a multi layer network using GRU and extracts local and global time series features via dilated connections. We learns attention mechanism to weight time series features and classify passive sonar signals. In experiments using public underwater acoustic data, the proposed network showed superior classification accuracy of 96.50 %. This result is 4.17 % higher classification accuracy compared to existing skip connected GRU network.

• Journal of the Korea Society of Computer and Information
• /
• v.28 no.4
• /
• pp.31-39
• /
• 2023

In this paper, we propose an Internal/External Knowledge Distillation (IEKD), which utilizes both external correlations between feature maps of heterogeneous models and internal correlations between feature maps of the same model for transferring knowledge from a teacher model to a student model. To achieve this, we transform feature maps into a sequence format and extract new feature maps suitable for knowledge distillation by considering internal and external correlations through a transformer. We can learn both internal and external correlations by distilling the extracted feature maps and improve the accuracy of the student model by utilizing the extracted feature maps with feature matching. To demonstrate the effectiveness of our proposed knowledge distillation method, we achieved 76.23% Top-1 image classification accuracy on the CIFAR-100 dataset with the "ResNet-32×4/VGG-8" teacher and student combination and outperformed the state-of-the-art KD methods.

• The Journal of the Acoustical Society of Korea
• /
• v.42 no.6
• /
• pp.500-510
• /
• 2023

Passive sonar signals mainly contain both normal and abnormal signals. The abnormal signals mixed with normal signals are primarily detected using an AutoEncoder (AE) that learns only normal signals. However, existing AEs may perform inaccurate detection by reconstructing distorted normal signals from mixed signal. To address these limitations, we propose an abnormal signal detection model based on a Recurrent Neural Network (RNN) and vector quantization. The proposed model generates a codebook representing the learned latent vectors and detects abnormal signals more accurately through the proposed search process of code vectors. In experiments using publicly available underwater acoustic data, the AE and Variational AutoEncoder (VAE) using the proposed method showed at least a 2.4 % improvement in the detection performance and at least a 9.2 % improvement in the extraction performance for abnormal signals than the existing models.

• Journal of the Institute of Convergence Signal Processing
• /
• v.24 no.1
• /
• pp.76-81
• /
• 2023

To distinguish cancer cells from normal cells, H&E (Hematoxylin & Eosin) staining is required. Pathological staining requires a lot of money and time. Recently, a digital dyeing method has been introduced to reduce such cost and time. In this paper, we propose a novel digital pathology algorithms. The first algorithm is the Pair method. This method learns the dyed phase image and unstained amplitude image taken by FPM (Fourier Ptychographic Microscopy) and converts it into a dyed amplitude image. The second algorithm is the unpair method. This method use the stained and unstained fluorescence microscopic images for modeling. In this study, digital staining was performed using a generative adversarial network (GAN). From the experimental results, we noticed that both the pair and unpair algorithms shows the excellent performance.

• Annual Conference on Human and Language Technology
• /
• 2003.10d
• /
• pp.54-60
• /
• 2003

음차 표기된 외래어로부터 원어를 복원하는 문제는 원어의 발음정보를 이용한 통계적인 방법을 많이 사용한다. 하지만 지금까지의 연구들은 대부분 영어단어만을 그 대상으로 하였기 때문에 '도쿄(Tokyo)', '하인리히(Hinrich)'와 같이 어원이 영어가 아닌 단어들의 복원에는 좋은 결과를 보여주지 못했다. 이러한 문제를 해결하기 위하여 한글로 표기된 외래어의 어원을 판단할 수 있는 방법을 찾아내고, 이 방법을 통해 외래어를 어원별로 분리하여 학습모델을 구축함으로써 다양한 어원을 가진 외래어들의 복원 정확률을 높이고자 하였다. 위의 방식으로 구현된 시스템은 영어, 일본어, 중국어, 프랑스어의, 서로 다른 4개의 어원을 가진 데이터의 복원 실험에서 기존의 방식에 비해 13%의 성능 향상을 보였다.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2023.05a
• /
• pp.226-226
• /
• 2023

하천유량, 댐유입량 등을 예측하기 위해 다양한 Long Short-Term Memory (LSTM) 방법들이 활발하게 적용 및 개발되고 있다. 최근 연구들은 s2s (sequence-to-sequence), Attention 기법 등을 통해 LSTM의 성능을 개선할 수 있음을 제시하고 있다. 이에 따라 본 연구에서는 LSTM-s2s와 LSTM-s2s에 attention까지 첨가한 모델을 구축하고, 시간 단위 자료를 사용하여 유입량 예측을 수행하여, 이의 실제 댐 운영에 모델들의 활용 가능성을 확인하고자 하였다. 소양강댐 유역을 대상으로 2013년부터 2020년까지의 유입량 시자료와 종관기상관측기온 및 강수량 데이터를 학습, 검증, 평가로 나누어 훈련한 후, 모델의 성능 평가를 진행하였다. 최적 시퀀스 길이를 결정하기 위해 R², RRMSE, CC, NSE, 그리고 PBIAS을 사용하였다. 분석 결과, LSTM-s2s 모델보다 attention까지 첨가한 모델이 전반적으로 성능이 우수했으며, attention 첨가 모델이 첨두값 예측에서도 높은 정확도를 보였다. 두 모델 모두 첨두값 발생 동안 유량 패턴을 잘 반영하였지만 세밀한 시간 단위 변화량 패턴 모의에는 한계가 있었다. 시간 단위 예측의 한계에도 불구하고, LSTM-s2s에 attention까지 추가한 모델은 향후 댐유입량 예측에 활용될 수 있을 것으로 판단한다.