• The Bulletin of The Korean Astronomical Society
• /
• v.46 no.2
• /
• pp.74.2-75
• /
• 2021

We demonstrate that a deep learning classifier that only uses to gravitational wave (GW) detectors auxiliary channel data can distinguish various types of non-Gaussian noise transients (glitches) with significant accuracy, i.e., ≳ 80%. The classifier is implemented using the multi-scale neural networks (MSNN) with PyTorch. The glitches appearing in the GW strain data have been one of the main obstacles that degrade the sensitivity of the gravitational detectors, consequently hindering the detection and parameterization of the GW signals. Numerous efforts have been devoted to tracking down their origins and to mitigating them. However, there remain many glitches of which origins are not unveiled. We apply the MSNN classifier to the auxiliary channel data corresponding to publicly available GravitySpy glitch samples of LIGO O1 run without using GW strain data. Investigation of the auxiliary channel data of the segments that coincide to the glitches in the GW strain channel is particularly useful for finding the noise sources, because they record physical and environmental conditions and the status of each part of the detector. By only using the auxiliary channel data, this classifier can provide us with the independent view on the data quality and potentially gives us hints to the origins of the glitches, when using the explainable AI technique such as Layer-wise Relevance Propagation or GradCAM.

• Journal of IKEEE
• /
• v.26 no.2
• /
• pp.176-185
• /
• 2022

Data augmentation is general approach to solve overfitting of machine learning models by applying various data transformations and distortions to dataset. However, when data augmentation is applied in GAN-based model, which is deep learning image generation model, data transformation and distortion are reflected in the generated image, then the generated image quality decrease. To prevent this problem called augmentation leak, we propose a scheme that can prevent augmentation leak regardless of the type and number of augmentations. Specifically, we analyze the conditions of augmentation leak occurrence by type and implement auxiliary augmentation task classifier that can prevent augmentation leak. Through experiments, we show that the proposed technique prevents augmentation leak in the GAN model, and as a result improves the quality of the generated image. We also demonstrate the superiority of the proposed scheme through ablation study and comparison with other representative augmentation leak prevention technique.

• Journal of Advanced Marine Engineering and Technology
• /
• v.28 no.5
• /
• pp.773-781
• /
• 2004

This research was Performed to raise grade efficiency of Sturtevant type air classifier. to treat powder less than $74\mu\textrm{m}$ particle produced at the crushing process of the dry aggregates manufacturing system or concrete wastes recycling system. The experimental conditions were in the ranges. 0.85 to 5.15 $m^3$/s of primary air flow rate. 0.005 to 0.015 $m^3$/s of secondary air flow rate $30^{\circ}$ to $70^{\circ}$ of auxiliary blades angle. respectively. for 1.7~3.3 kg/min of the powder feed rate. It was found that the grade efficiency of the air classifier was increased as the baffle plate was attached at the expansion region. and the optimal operating conditions of the air flow rates and the blade angle were obtained. The fractional recovery curves from the experiments were well agreement with the theoretical one of Molerus model.

• Journal of Korea Society of Digital Industry and Information Management
• /
• v.19 no.1
• /
• pp.91-107
• /
• 2023

With the recent expansion of the commerce ecosystem, a large number of user evaluations have been produced. Accordingly, attempts to create business insights using user evaluation data have been actively made. However, since user evaluation can change after the user experiences the product, it is difficult to say that the analysis based only on reviews immediately after purchase fully reflects the user's evaluation of the product. Moreover, studies conducted so far on user evaluation have overlooked the fact that the length of time a user has used a product can affect the user's product evaluation. Therefore, in this study, we build a model that predicts the direction of change in the user's rating after use from the user's rating and reviews immediately after purchase. In particular, the proposed model reflects the product's period of use in predicting the change direction of the star rating. However, since the posterior information on the duration of product use cannot be used as input in the inference process, we propose a structure that utilizes information about the product's period of use using an auxiliary classifier. As a result of an experiment using 599,889 user evaluation data collected from the shopping platform 'N' company, we confirmed that the proposed model performed better than the existing model in terms of accuracy.

• Journal of Intelligence and Information Systems
• /
• v.28 no.3
• /
• pp.119-137
• /
• 2022

Recent recommendation system studies apply various deep learning models to represent user and item interactions better. One of the noteworthy studies is ONCF(Outer product-based Neural Collaborative Filtering) which builds a two-dimensional interaction map via outer product and employs CNN (Convolutional Neural Networks) to learn high-order correlations from the map. However, ONCF has limitations in recommendation performance due to the problems with CNN and the absence of side information. ONCF using CNN has an inductive bias problem that causes poor performances for data with a distribution that does not appear in the training data. This paper proposes to employ a Vision Transformer (ViT) instead of the vanilla CNN used in ONCF. The reason is that ViT showed better results than state-of-the-art CNN in many image classification cases. In addition, we propose a new architecture to reflect side information that ONCF did not consider. Unlike previous studies that reflect side information in a neural network using simple input combination methods, this study uses an independent auxiliary classifier to reflect side information more effectively in the recommender system. ONCF used a single latent vector for user and item, but in this study, a channel is constructed using multiple vectors to enable the model to learn more diverse expressions and to obtain an ensemble effect. The experiments showed our deep learning model improved performance in recommendation compared to ONCF.

• Journal of Electrical Engineering and Technology
• /
• v.11 no.6
• /
• pp.1872-1879
• /
• 2016

The techniques to recycle and reuse plastics attract public attention. These public attraction and needs result in improving the recycling technique. However, the identification technique for black plastic wastes still have big problem that the spectrum extracted from near infrared radiation spectroscopy is not clear and is contaminated by noise. To overcome this problem, we apply Raman spectroscopy to extract a clear spectrum of plastic material. In addition, to improve the classification ability of fuzzy Radial Basis Function Neural Networks, we apply supervised learning based clustering method instead of unsupervised clustering method. The conditional fuzzy C-Means clustering method, which is a kind of supervised learning based clustering algorithms, is used to determine the location of radial basis functions. The conditional fuzzy C-Means clustering analyzes the data distribution over input space under the supervision of auxiliary information. The auxiliary information is defined by using k Nearest Neighbor approach.

• Smart Media Journal
• /
• v.8 no.4
• /
• pp.33-37
• /
• 2019

Because the ECG signals are time-series data acquired as time elapses, it is important to obtain comparative data the same in size as the enrolled data every time. This paper suggests a network model of GAN (Generative Adversarial Networks) based on an auxiliary classifier to generate synthetic ECG signals which may address the different data size issues. The Cosine similarity and Cross-correlation are used to examine the similarity of synthetic ECG signals. The analysis shows that the Average Cosine similarity was 0.991 and the Average Euclidean distance similarity based on cross-correlation was 0.25: such results indicate that data size difference issue can be resolved while the generated synthetic ECG signals, similar to real ECG signals, can create synthetic data even when the registered data are not the same as the comparative data in size.

• Journal of the Korea Society of Computer and Information
• /
• v.29 no.3
• /
• pp.11-20
• /
• 2024

Among the generative models in Artificial Intelligence (AI), especially Generative Adversarial Network (GAN) has been successful in various applications such as image processing, density estimation, and style transfer. While the GAN models including Conditional GAN (CGAN), CycleGAN, BigGAN, have been extended and improved, researchers face challenges in real-world applications in specific domains such as disaster simulation, healthcare, and urban planning due to data scarcity and unstable learning causing Image distortion. This paper proposes a new progressive learning methodology called Progressive Step Training (PST) based on the Auxiliary Classifier GAN (ACGAN) that discriminates class labels, leveraging the progressive learning approach of the Progressive Growing of GAN (PGGAN). The PST model achieves 70.82% faster stabilization, 51.3% lower standard deviation, stable convergence of loss values in the later high resolution stages, and a 94.6% faster loss reduction compared to conventional methods.

• Journal of Intelligence and Information Systems
• /
• v.28 no.3
• /
• pp.185-207
• /
• 2022

Recently, with the development of deep learning technology, researches to apply a deep learning algorithm to analyze unstructured data such as text and images are being actively conducted. Text classification has been studied for a long time in academia and industry, and various attempts are being performed to utilize data characteristics to improve classification performance. In particular, a hierarchical relationship of labels has been utilized for hierarchical classification. However, the top-down approach mainly used for hierarchical classification has a limitation that misclassification at a higher level blocks the opportunity for correct classification at a lower level. Therefore, in this study, we propose a methodology for classifying hierarchical data using the autoencoder-based deeply supervised network that high-level classification does not block the low-level classification while considering the hierarchical relationship of labels. The proposed methodology adds a main classifier that predicts a low-level label to the autoencoder's latent variable and an auxiliary classifier that predicts a high-level label to the hidden layer of the autoencoder. As a result of experiments on 22,512 academic papers to evaluate the performance of the proposed methodology, it was confirmed that the proposed model showed superior classification accuracy and F1-score compared to the traditional supervised autoencoder and DNN model.

• The Bulletin of The Korean Astronomical Society
• /
• v.36 no.2
• /
• pp.131.2-131.2
• /
• 2011

We present performance of artificial neural network multivariate classifier in identifying non-astrophysical origin noise transients from the gravitational wave channel of Laser Interferometer Gravitational-wave Observatory (LIGO). LIGO has successfully conducted six science runs, achieving the sensitivity as planned and producing many fruitful scientific results. It has been well observed that the detector noise is non-Gaussian and non-stationary, which results in large excess of noise transients called glitches arising from instrumental and environmental artifacts. Great efforts have been committed to reduce the glitches by tuning the detector instruments and by vetoing them but further improvement is still needed. To this end, there have been efforts to incorporate data from hundreds of auxiliary, physical and environmental channels into identifying the glitches in the gravitational wave channel. We introduce a multivariate classification method using Artificial Neural Networks (ANNs) that efficiently handles large number of variables. In this poster, we present preliminary results of the application of our ANN algorithm to data from LIGO's Science Run 4 and compare its performance with conventional vetoing method.