• Journal of The Korean Astronomical Society
• /
• v.56 no.2
• /
• pp.213-224
• /
• 2023

Type II solar radio bursts show frequency drifts from high to low over time. They have been known as a signature of coronal shock associated with Coronal Mass Ejections (CMEs) and/or flares, which cause an abrupt change in the space environment near the Earth (space weather). Therefore, early detection of type II bursts is important for forecasting of space weather. In this study, we develop a deep-learning (DL) model for the automatic detection of type II bursts. For this purpose, we adopted a 1-D Convolution Neutral Network (CNN) as it is well-suited for processing spatiotemporal information within the applied data set. We utilized a total of 286 radio burst spectrum images obtained by Hiraiso Radio Spectrograph (HiRAS) from 1991 and 2012, along with 231 spectrum images without the bursts from 2009 to 2015, to recognizes type II bursts. The burst types were labeled manually according to their spectra features in an answer table. Subsequently, we applied the 1-D CNN technique to the spectrum images using two filter windows with different size along time axis. To develop the DL model, we randomly selected 412 spectrum images (80%) for training and validation. The train history shows that both train and validation losses drop rapidly, while train and validation accuracies increased within approximately 100 epoches. For evaluation of the model's performance, we used 105 test images (20%) and employed a contingence table. It is found that false alarm ratio (FAR) and critical success index (CSI) were 0.14 and 0.83, respectively. Furthermore, we confirmed above result by adopting five-fold cross-validation method, in which we re-sampled five groups randomly. The estimated mean FAR and CSI of the five groups were 0.05 and 0.87, respectively. For experimental purposes, we applied our proposed model to 85 HiRAS type II radio bursts listed in the NGDC catalogue from 2009 to 2016 and 184 quiet (no bursts) spectrum images before and after the type II bursts. As a result, our model successfully detected 79 events (93%) of type II events. This results demonstrates, for the first time, that the 1-D CNN algorithm is useful for detecting type II bursts.

• Journal of Internet Computing and Services
• /
• v.25 no.1
• /
• pp.109-121
• /
• 2024

In various manufacturing processes such as textiles and automobiles, when equipment breaks down or stops, the machines do not work, which leads to time and financial losses for the company. Therefore, it is important to detect equipment abnormalities in advance so that equipment failures can be predicted and repaired before they occur. Most equipment failures are caused by bearing failures, which are essential parts of equipment, and detection bearing anomaly is the essence of PHM(Prognostics and Health Management) research. In this paper, we propose a preprocessing algorithm called SWT-SVD, which analyzes vibration signals from bearings and apply it to an anomaly transformer, one of the time series anomaly detection model networks, to implement bearing anomaly detection model. Vibration signals from the bearing manufacturing process contain noise due to the real-time generation of sensor values. To reduce noise in vibration signals, we use the Stationary Wavelet Transform to extract frequency components and perform preprocessing to extract meaningful features through the Singular Value Decomposition algorithm. For experimental validation of the proposed SWT-SVD preprocessing method in the bearing anomaly detection model, we utilize the PHM-2012-Challenge dataset provided by the IEEE PHM Conference. The experimental results demonstrate significant performance with an accuracy of 0.98 and an F1-Score of 0.97. Additionally, to substantiate performance improvement, we conduct a comparative analysis with previous studies, confirming that the proposed preprocessing method outperforms previous preprocessing methods in terms of performance.

• The Plant Pathology Journal
• /
• v.40 no.1
• /
• pp.48-58
• /
• 2024

The oldest and most extensively cultivated form of millet, known as pearl millet (Pennisetum glaucum (L.) R. Br. Syn. Pennisetum americanum (L.) Leeke), is raised over 312.00 lakh hectares in Asian and African countries. India is regarded as the significant hotspot for pearl millet diversity. In the Indian state of Haryana, where pearl millet is grown, a new and catastrophic bacterial disease known as stem rot of pearl millet spurred by the bacterium Klebsiella aerogenes (formerly Enterobacter) was first observed during fall 2018. The disease appears in form of small to long streaks on leaves, lesions on stem, and slimy rot appearance of stem. The associated bacterium showed close resemblance to Klebsiella aerogenes that was confirmed by a molecular evaluation based on 16S rDNA and gyrA gene nucleotide sequences. The isolates were also identified to be Klebsiella aerogenes based on biochemical assays, where Klebsiella isolates differed in D-trehalose and succinate alkalisation tests. During fall 2021-2023, the disease has spread all the pearl millet-growing districts of the state, extending up to 70% disease incidence in the affected fields. The disease is causing considering grain as well as fodder losses. The proposed scale, consisting of six levels (0-5), is developed where scores 0, 1, 2, 3, 4, and 5 have been categorized as highly resistant, resistant, moderately resistant, moderately susceptible, susceptible, and highly susceptible disease reaction, respectively. The disease cycle, survival of pathogen, and possible losses have also been studied to understand other features of the disease.

• Journal of Korean Society of Forest Science
• /
• v.113 no.1
• /
• pp.14-30
• /
• 2024

This study characterized areas at risk of land creep by focusing on a site that has undergone this phenomenon in Ulju-gun, South Korea. Land creep in the area of interest was catalyzed by road expansion work conducted in 2022. The site was examined on the basis of its geological features, topography, effective soil depth, soil hardness, electrical resistivity, and subsurface profile. It consists of a slope covered with sparse vegetation and a concave top that retains rainwater during rainfall. Compositionally, land creep affected the shale, sandstone, and conglomerate formations on the site, which had little silt and more sand and clay compared with areas that were unaffected by land creep. An electrical resistivity survey enabled us to detect a groundwater zone at the site, which explains the softness of the soil. Finally, the effective soil depth at the land creep-affected area was 30.4 cm on average, indicating deep colluvial deposits. In contrast, unaffected sites had an effective soil depth of 24.7 cm on average. These results should facilitate the creation of systems for monitoring and preemptively responding to land creep, significantly mitigating the socioeconomic losses associated with this phenomenon.

• Journal of Intelligence and Information Systems
• /
• v.24 no.3
• /
• pp.1-19
• /
• 2018

Large amount of data is now available for research and business sectors to extract knowledge from it. This data can be in the form of unstructured data such as audio, text, and image data and can be analyzed by deep learning methodology. Deep learning is now widely used for various estimation, classification, and prediction problems. Especially, fashion business adopts deep learning techniques for apparel recognition, apparel search and retrieval engine, and automatic product recommendation. The core model of these applications is the image classification using Convolutional Neural Networks (CNN). CNN is made up of neurons which learn parameters such as weights while inputs come through and reach outputs. CNN has layer structure which is best suited for image classification as it is comprised of convolutional layer for generating feature maps, pooling layer for reducing the dimensionality of feature maps, and fully-connected layer for classifying the extracted features. However, most of the classification models have been trained using online product image, which is taken under controlled situation such as apparel image itself or professional model wearing apparel. This image may not be an effective way to train the classification model considering the situation when one might want to classify street fashion image or walking image, which is taken in uncontrolled situation and involves people's movement and unexpected pose. Therefore, we propose to train the model with runway apparel image dataset which captures mobility. This will allow the classification model to be trained with far more variable data and enhance the adaptation with diverse query image. To achieve both convergence and generalization of the model, we apply Transfer Learning on our training network. As Transfer Learning in CNN is composed of pre-training and fine-tuning stages, we divide the training step into two. First, we pre-train our architecture with large-scale dataset, ImageNet dataset, which consists of 1.2 million images with 1000 categories including animals, plants, activities, materials, instrumentations, scenes, and foods. We use GoogLeNet for our main architecture as it has achieved great accuracy with efficiency in ImageNet Large Scale Visual Recognition Challenge (ILSVRC). Second, we fine-tune the network with our own runway image dataset. For the runway image dataset, we could not find any previously and publicly made dataset, so we collect the dataset from Google Image Search attaining 2426 images of 32 major fashion brands including Anna Molinari, Balenciaga, Balmain, Brioni, Burberry, Celine, Chanel, Chloe, Christian Dior, Cividini, Dolce and Gabbana, Emilio Pucci, Ermenegildo, Fendi, Giuliana Teso, Gucci, Issey Miyake, Kenzo, Leonard, Louis Vuitton, Marc Jacobs, Marni, Max Mara, Missoni, Moschino, Ralph Lauren, Roberto Cavalli, Sonia Rykiel, Stella McCartney, Valentino, Versace, and Yve Saint Laurent. We perform 10-folded experiments to consider the random generation of training data, and our proposed model has achieved accuracy of 67.2% on final test. Our research suggests several advantages over previous related studies as to our best knowledge, there haven't been any previous studies which trained the network for apparel image classification based on runway image dataset. We suggest the idea of training model with image capturing all the possible postures, which is denoted as mobility, by using our own runway apparel image dataset. Moreover, by applying Transfer Learning and using checkpoint and parameters provided by Tensorflow Slim, we could save time spent on training the classification model as taking 6 minutes per experiment to train the classifier. This model can be used in many business applications where the query image can be runway image, product image, or street fashion image. To be specific, runway query image can be used for mobile application service during fashion week to facilitate brand search, street style query image can be classified during fashion editorial task to classify and label the brand or style, and website query image can be processed by e-commerce multi-complex service providing item information or recommending similar item.

• Investigative Magnetic Resonance Imaging
• /
• v.10 no.1
• /
• pp.8-15
• /
• 2006

Purpose : To describe MR imaging features of hypoxic brain damage in relation to time elapse and prognosis of patients. Materials and methods : We reviewed 19 MR studies of 18 patients with hypoxic brain damage. MR imaging studies were performed between 1 to 20 days after the hypoxic insults (mean 8.6 days). MR images were analyzed with regard to the locations of abnormal signal intensities, the presence of brain edema. And imaging findings were correlated with the time elapse after the insults and the prognosis of patients. Results : On 19 cases of MR studies, abnormal high intensities on T2-weighted images were found in the basal ganglia (15, 78.9%), cerebral cortex (13, 68.4%), white matter (9, 47.4%), thalamus (6, 31.6%), cerebellum (4, 21.1%) and brainstem (1, 5.3%), respectively. Cerebral cortical involvement was typically bilateral and diffuse, but sometimes limited to the parieto-occipital area. The brainstem and cerebellar involvement was rare and in all cases, cerebral cortical lesions accompanied. Most of the white matter lesions were accompanied with cortical and deep gray matter lesions and found in subacute period(>6 days). The cortical high signal intensity lesions on T1-weighted image were found mostly in subacute stage, but in some cases involvement was also found in acute stage ($\leq$ 6 days). The cortical edema is found on 11 cases in acute and subacute stages. In cases of recovered consciousness, cortical involvement and edema on MR were rare. Conclusion : MR findings of hypoxic brain damage were various, but diffuse bilateral involvement of cortex and/or deep gray matter was found in most of the cases. White matter involvement was rarely found in acute stage and usually found in subacute stage. In cases of good pronosis, cortical involvement and edema were rare.

• KOREAN JOURNAL OF CROP SCIENCE
• /
• v.38 no.5
• /
• pp.405-412
• /
• 1993

This experiment was conducted to investigate the effects of water depths on seedling stand and early growth of califonia rice varieties, S201, M202, A301, Italico livorno and Korean variety, Hwaseongbyeo, and barnyardgrass (Echinochloa crus-galli) The coleoptile length of rice was longer with deep water depth while for the radicle length shorten. As water depth was increased, the percentage of seedling stand were decreased slightly in rice, while sharply increased in barnyardgrass. Plant height of rice with increasing water depth were longer, whereas that of barnyardgrass reduced significantly with weaker. Tiller number of rice and barnyardgrass were significantly reduced as water depth increased. Dry matter weight and healthy score of rice seedling at 35DAS were highest in 7.5cm water depth followed saturated moisture, 15, and 22.5cm water depth, while for barnyardgrass those were especially negatively affected by deep water depth. These results showed that the seedling stand and early growth of barnyardgrass was highly suppressed by deeper water levels compared with rice. Rice cultivars which are showes growth characteristics in deeper water levels at early growth stage were Italico livorno and S201 in Japonica / Indica.

• Journal of Intelligence and Information Systems
• /
• v.25 no.4
• /
• pp.105-122
• /
• 2019

Dimensionality reduction is one of the methods to handle big data in text mining. For dimensionality reduction, we should consider the density of data, which has a significant influence on the performance of sentence classification. It requires lots of computations for data of higher dimensions. Eventually, it can cause lots of computational cost and overfitting in the model. Thus, the dimension reduction process is necessary to improve the performance of the model. Diverse methods have been proposed from only lessening the noise of data like misspelling or informal text to including semantic and syntactic information. On top of it, the expression and selection of the text features have impacts on the performance of the classifier for sentence classification, which is one of the fields of Natural Language Processing. The common goal of dimension reduction is to find latent space that is representative of raw data from observation space. Existing methods utilize various algorithms for dimensionality reduction, such as feature extraction and feature selection. In addition to these algorithms, word embeddings, learning low-dimensional vector space representations of words, that can capture semantic and syntactic information from data are also utilized. For improving performance, recent studies have suggested methods that the word dictionary is modified according to the positive and negative score of pre-defined words. The basic idea of this study is that similar words have similar vector representations. Once the feature selection algorithm selects the words that are not important, we thought the words that are similar to the selected words also have no impacts on sentence classification. This study proposes two ways to achieve more accurate classification that conduct selective word elimination under specific regulations and construct word embedding based on Word2Vec embedding. To select words having low importance from the text, we use information gain algorithm to measure the importance and cosine similarity to search for similar words. First, we eliminate words that have comparatively low information gain values from the raw text and form word embedding. Second, we select words additionally that are similar to the words that have a low level of information gain values and make word embedding. In the end, these filtered text and word embedding apply to the deep learning models; Convolutional Neural Network and Attention-Based Bidirectional LSTM. This study uses customer reviews on Kindle in Amazon.com, IMDB, and Yelp as datasets, and classify each data using the deep learning models. The reviews got more than five helpful votes, and the ratio of helpful votes was over 70% classified as helpful reviews. Also, Yelp only shows the number of helpful votes. We extracted 100,000 reviews which got more than five helpful votes using a random sampling method among 750,000 reviews. The minimal preprocessing was executed to each dataset, such as removing numbers and special characters from text data. To evaluate the proposed methods, we compared the performances of Word2Vec and GloVe word embeddings, which used all the words. We showed that one of the proposed methods is better than the embeddings with all the words. By removing unimportant words, we can get better performance. However, if we removed too many words, it showed that the performance was lowered. For future research, it is required to consider diverse ways of preprocessing and the in-depth analysis for the co-occurrence of words to measure similarity values among words. Also, we only applied the proposed method with Word2Vec. Other embedding methods such as GloVe, fastText, ELMo can be applied with the proposed methods, and it is possible to identify the possible combinations between word embedding methods and elimination methods.

• Korean Journal of Heritage: History & Science
• /
• v.54 no.1
• /
• pp.52-69
• /
• 2021

The transcriptions of Buddhist sutra in the Goryeo Dynasty are more elaborate and splendid than those of any other period and occupy a very important position in Korean bibliography. Among them, the transcriptions made on indigo paper show decorative features that represent the dignity and quality that nobles would have preferred. Particularly, during the Goryeo Dynasty, a large number of transcriptions were made on indigo paper, often in hand-scrolled and folded forms. If flexibility was not guaranteed, the hand-scrolled form caused inconvenience and damage when handling the transcription because of the structural limitations of the material that is rolled up and opened. It was possible to overcome these shortcomings by changing from the hand-scrolled to the folded form to obtain convenience and structural stability. The folded form of the transcription utilizes the same principle as the folding screen, so it is a structure that can be folded and unfolded, and it is made by connecting parts at regularly spaced intervals. No matter how small the transcription is, if it is made of thin paper, it is difficult to handle it and to maintain its shape and structure. For this reason, the folded transcription was usually made of thick paper to support the structure, and the cover was made thicker than the inner part to protect the contents. In other words, the forded form was generally manufactured to suit the characteristics of maintaining strength by making the paper thick. Because a large amount of indigo paper was needed to make this type of transcription, it is assumed that there were craftsmen who were in charge only of dark dyeing the papers. Usually, paper dyeing requires much more dye than silk dyeing, and dyeing dozens of times would be required to obtain the deep indigo color of the base of the transcription of Buddhist sutra in the Goryeo Dynasty. Unfortunately, there is no record of the Goryeo Dynasty's indigo blue paper manufacturing technique, and the craftsmen who made indigo paper no longer remain, so no one knows the exact method of making indigo paper. Recently, Hanji artisans, natural dyers, and conservators attempted to restore the Goryeo Dynasty's indigo paper, but the texture and deep colors found in the relics could not be reproduced. This study introduces the process of restoring indigo paper in the Goryeo Dynasty through collaboration between dyeing artisans, Hanji artisans, and conservators for conservation of the transcription of Buddhist sutra in the late Goryeo dynasty, yielding a suggested method of making indigo paper.

• Journal of Popular Narrative
• /
• v.26 no.3
• /
• pp.189-219
• /
• 2020

This study is an analysis of Handol Heung-Gun Lee's Tarae, which is a coinage combining the Korean words for "playing an instrument" and "song", in terms of narrative and aesthetics. The components for analysis are the phenomena and nature of binary oppositions between nature and human beings, between alienation and interest, between division and unification, and between diaspora and people of the national community. Tarae in the period from the late 1970s to the early 1990s described the experience of pain and loss from non-resistance and disobedience in protest against social problems that emerged during the era of miliary dictatorship, such as industrialization, urbanization, reckless development, Westernization, university-oriented education, the gap between rich and poor, human alienation, and the conflicts arising from the division of the nation. After Handol overcame the lack of creative motivation with self-reflection and effort, Tarae took the form of a diaspora epic meta-narratives integrating the "sound of nature and his true nature" and "the awareness of diaspora and the spirit of the Korean people". The epics of the homeland, the national soil and the people, which began with "Teo", became more intense in terms of a sense of diaspora as they shifted their focus from an origin to a path with "Hanmoejulghi" as the turning point. Handol seeks inspiration in the source of narrative rather than in music. His Tarae focuses on "adding rhythm for lyrics". For this reason, the semiotic features of Tarae have a limitation in that its extrinsic phonology is simple even if its intrinsic meaning (i.e., emotion of sadness) is profound and subtle. In order to elicit sympathy from the audience and impress them, it is necessary to strike a balance between the implicit (semantic) part and the explicit (phonological) part. To share the emotion of sadness with more people, it is necessary to strengthen phonological elements. Sympathy for sadness and deep impression on the audience are more often induced by the mood of similar sentiments than by the stories of the same experience. The aesthetics of sadness in Tarae began with the narratives of past experience which were expressed in the contexts of loss, loneliness, and poverty that Handol had experienced since childhood. However, the aesthetics of sadness, deepened over the period of a long hiatus in Handol's career as a composer, formed the narratives of ultimate salvation, embodying even the diaspora experience of others (e.g., displaced people, overseas adoptees, ethnic Koreans in Russia, victims of Japanese military sexual slavery, etc.). This gave Tarae the potential to go beyond the limits of the ethnic group of Korea. Tarae, as a "dispersed sound", can benefit from the appeal of deep sadness at the point of contact with other forms of world music. It may form a global diaspora discourse because Tarae is oriented towards interculturalism rather than anti-multiculturalism. The future challenge and goal of Handol's Tarae would be to continue to find areas of sympathy and broaden the horizon of awareness as diaspora music.