• Asia pacific journal of information systems
• /
• v.19 no.2
• /
• pp.157-178
• /
• 2009

Corporate credit rating is a very important factor in the market for corporate debt. Information concerning corporate operations is often disseminated to market participants through the changes in credit ratings that are published by professional rating agencies, such as Standard and Poor's (S&P) and Moody's Investor Service. Since these agencies generally require a large fee for the service, and the periodically provided ratings sometimes do not reflect the default risk of the company at the time, it may be advantageous for bond-market participants to be able to classify credit ratings before the agencies actually publish them. As a result, it is very important for companies (especially, financial companies) to develop a proper model of credit rating. From a technical perspective, the credit rating constitutes a typical, multiclass, classification problem because rating agencies generally have ten or more categories of ratings. For example, S&P's ratings range from AAA for the highest-quality bonds to D for the lowest-quality bonds. The professional rating agencies emphasize the importance of analysts' subjective judgments in the determination of credit ratings. However, in practice, a mathematical model that uses the financial variables of companies plays an important role in determining credit ratings, since it is convenient to apply and cost efficient. These financial variables include the ratios that represent a company's leverage status, liquidity status, and profitability status. Several statistical and artificial intelligence (AI) techniques have been applied as tools for predicting credit ratings. Among them, artificial neural networks are most prevalent in the area of finance because of their broad applicability to many business problems and their preeminent ability to adapt. However, artificial neural networks also have many defects, including the difficulty in determining the values of the control parameters and the number of processing elements in the layer as well as the risk of over-fitting. Of late, because of their robustness and high accuracy, support vector machines (SVMs) have become popular as a solution for problems with generating accurate prediction. An SVM's solution may be globally optimal because SVMs seek to minimize structural risk. On the other hand, artificial neural network models may tend to find locally optimal solutions because they seek to minimize empirical risk. In addition, no parameters need to be tuned in SVMs, barring the upper bound for non-separable cases in linear SVMs. Since SVMs were originally devised for binary classification, however they are not intrinsically geared for multiclass classifications as in credit ratings. Thus, researchers have tried to extend the original SVM to multiclass classification. Hitherto, a variety of techniques to extend standard SVMs to multiclass SVMs (MSVMs) has been proposed in the literature Only a few types of MSVM are, however, tested using prior studies that apply MSVMs to credit ratings studies. In this study, we examined six different techniques of MSVMs: (1) One-Against-One, (2) One-Against-AIL (3) DAGSVM, (4) ECOC, (5) Method of Weston and Watkins, and (6) Method of Crammer and Singer. In addition, we examined the prediction accuracy of some modified version of conventional MSVM techniques. To find the most appropriate technique of MSVMs for corporate bond rating, we applied all the techniques of MSVMs to a real-world case of credit rating in Korea. The best application is in corporate bond rating, which is the most frequently studied area of credit rating for specific debt issues or other financial obligations. For our study the research data were collected from National Information and Credit Evaluation, Inc., a major bond-rating company in Korea. The data set is comprised of the bond-ratings for the year 2002 and various financial variables for 1,295 companies from the manufacturing industry in Korea. We compared the results of these techniques with one another, and with those of traditional methods for credit ratings, such as multiple discriminant analysis (MDA), multinomial logistic regression (MLOGIT), and artificial neural networks (ANNs). As a result, we found that DAGSVM with an ordered list was the best approach for the prediction of bond rating. In addition, we found that the modified version of ECOC approach can yield higher prediction accuracy for the cases showing clear patterns.

• Information Systems Review
• /
• v.16 no.3
• /
• pp.49-68
• /
• 2014

These days, consumers have increasingly preferred to digital real-time streamlining and downloading to listen to music because this is convenient and affordable for the consumers. Accordingly, sales of music in compact disk formats have steadily declined. In this regards, online digital music has become a new communication channel to listen musics, where digital files can be delivered over various online networks to people's computing devices. The majority of online digital music distributors has Music Recommender Systems for sales of digital music on their websites. Music Recommender Systems are parts of information filtering systems that provide the ratings or preferences that users give to music. Korean online digital music distributors have Music Recommender Systems. But those online music distributors didn't provide any rules or clear procedures that recommend music. Therefore, we raise important questions as follows: "Is Music Recommender Systems Fair?", "What is the impact of Music Recommender Systems on online music rankings and sales?" While previous studies have focused on usefulness of Music Recommender Systems, this study investigates not only fairness of Current Music Recommender Systems but also Relationship between Music Recommender Systems and online Music Charts. This study examines these issues based on Bandwagon effect, ranking effect, Slot effect theories. For our empirical analysis, we selected the most famous five online digital music distributors in terms of market shares. We found that all recommended music is exposed to the top of 'daily music charts' in online digital music distributors' websites. We collected music ranking data and recommended music data from 'daily music chart' during a one month. The result shows that online music recommender systems are not fair, since they mainly recommend particular music that supported by a specific music production company. In addition, the recommended music are always exposed to the top of music ranking charts. We also find that recommended music usually appear at the top 20 ranking charts within one or two days. Also, the most music in the top 50 or 100 ranks are the recommended music. Moreover, recommended music usually remain the ranking charts more than one month while non-recommended music often disappear at the ranking charts within two week. Our study provides an important implication to online music industry. Because music recommender systems and music ranking charts are closely related, music distributors may improperly use their recommender systems to boost the sales of music that related to their own companies. Therefore, online digital music distributor must clearly announce the rules and procedures about music recommender systems for the better music industry.

• Proceedings of the Korean DIstribution Association Conference
• /
• 2005.11a
• /
• pp.29-51
• /
• 2005

Brands are important in the consumer market. They are the interface between consumers and the company, consumers may develop loyalty to brands. also, The late development of industrial marketing explains the near absence of research on Brand Equity in business to business. With recent change, industrial companies have shifted from a production focus to a customer focus. industrial brand is fast developing. The basic purpose of this study is to investigate industrial brand trust and loyalty affecting the Result of business relationship between industrial buyers and suppliers. Factors hypothesized to influence trust in a brand include a number of brand characteristics, company characteristics and consumer-brand characteristics. This research presented a comprehensive constructive model consisting of components of industrial brand trust and loyalty, and then propose the research model base on prior researches and studies about relationships among components of industrial brand loyalty. Data were gathered from respondents who work in industrial buying center. For this study, Data were analyzed by SPSS 10.0 and AMOS 4.0. The results of this research analysis were as fallow. Industrial brand trust and loyalty were positively related with a number of industrial brand characteristics, supplier characteristics and buyer-brand characteristics. relationship commitment. This research newly proposed the concept of 'industrial brand trust and loyalty affecting the Result of business relationship between industrial buyers and suppliers'

• Korean Chemical Engineering Research
• /
• v.60 no.1
• /
• pp.86-92
• /
• 2022

Recently, the bio-healthcare market is enlarging worldwide due to various reasons such as the COVID-19 pandemic. Among them, biometric measurement and analysis technology are expected to bring about future technological innovation and socio-economic ripple effect. Existing systems require a large-capacity battery to drive signal processing, wireless transmission part, and an operating system in the process. However, due to the limitation of the battery capacity, it causes a spatio-temporal limitation on the use of the device. This limitation can act as a cause for the disconnection of data required for the user's health care monitoring, so it is one of the major obstacles of the health care device. In this study, we report the concept of a standalone healthcare monitoring module, which is based on both triboelectric effects and electromagnetic effects, by converting biomechanical energy into suitable electric energy. The proposed system can be operated independently without an external power source. In particular, the wireless foot pressure measurement monitoring system, which is rationally designed triboelectric sensor (TES), can recognize the user's walking habits through foot pressure measurement. By applying the triboelectric effects to the contact-separation behavior that occurs during walking, an effective foot pressure sensor was made, the performance of the sensor was verified through an electrical output signal according to the pressure, and its dynamic behavior is measured through a signal processing circuit using a capacitor. In addition, the biomechanical energy dissipated during walking is harvested as electrical energy by using the electromagnetic induction effect to be used as a power source for wireless transmission and signal processing. Therefore, the proposed system has a great potential to reduce the inconvenience of charging caused by limited battery capacity and to overcome the problem of data disconnection.

• Korean Journal of Remote Sensing
• /
• v.39 no.5_3
• /
• pp.979-995
• /
• 2023

As wildfires are difficult to predict, real-time monitoring is crucial for a timely response. Geostationary satellite images are very useful for active fire detection because they can monitor a vast area with high temporal resolution (e.g., 2 min). Existing satellite-based active fire detection algorithms detect thermal outliers using threshold values based on the statistical analysis of brightness temperature. However, the difficulty in establishing suitable thresholds for such threshold-based methods hinders their ability to detect fires with low intensity and achieve generalized performance. In light of these challenges, machine learning has emerged as a potential-solution. Until now, relatively simple techniques such as random forest, Vanilla convolutional neural network (CNN), and U-net have been applied for active fire detection. Therefore, this study proposed an active fire detection algorithm using state-of-the-art (SOTA) deep learning techniques using data from the Advanced Himawari Imager and evaluated it over East Asia and Australia. The SOTA model was developed by applying EfficientNet and lion optimizer, and the results were compared with the model using the Vanilla CNN structure. EfficientNet outperformed CNN with F1-scores of 0.88 and 0.83 in East Asia and Australia, respectively. The performance was better after using weighted loss, equal sampling, and image augmentation techniques to fix data imbalance issues compared to before the techniques were used, resulting in F1-scores of 0.92 in East Asia and 0.84 in Australia. It is anticipated that timely responses facilitated by the SOTA deep learning-based approach for active fire detection will effectively mitigate the damage caused by wildfires.

• Journal of Korea Water Resources Association
• /
• v.57 no.3
• /
• pp.151-164
• /
• 2024

In water treatment plants supplying potable water, the management of chlorine concentration in water treatment processes involving pre-chlorination or intermediate chlorination requires process control. To address this, research has been conducted on water quality prediction techniques utilizing AI technology. This study developed an AI-based predictive model for automating the process control of chlorine disinfection, targeting the prediction of residual chlorine concentration downstream of sedimentation basins in water treatment processes. The AI-based model, which learns from past water quality observation data to predict future water quality, offers a simpler and more efficient approach compared to complex physicochemical and biological water quality models. The model was tested by predicting the residual chlorine concentration downstream of the sedimentation basins at Plant, using multiple regression models and AI-based models like Random Forest and LSTM, and the results were compared. For optimal prediction of residual chlorine concentration, the input-output structure of the AI model included the residual chlorine concentration upstream of the sedimentation basin, turbidity, pH, water temperature, electrical conductivity, inflow of raw water, alkalinity, NH₃, etc. as independent variables, and the desired residual chlorine concentration of the effluent from the sedimentation basin as the dependent variable. The independent variables were selected from observable data at the water treatment plant, which are influential on the residual chlorine concentration downstream of the sedimentation basin. The analysis showed that, for Plant, the model based on Random Forest had the lowest error compared to multiple regression models, neural network models, model trees, and other Random Forest models. The optimal predicted residual chlorine concentration downstream of the sedimentation basin presented in this study is expected to enable real-time control of chlorine dosing in previous treatment stages, thereby enhancing water treatment efficiency and reducing chemical costs.

• Journal of Life Science
• /
• v.30 no.9
• /
• pp.804-812
• /
• 2020

The giant mealworm beetle, Zophobas atratus (Coleoptera: Tenebrionidae) has been used as a protein source for small pets and mammals. Recently, it was temporarily registered in the list of the Food Code. We previously performed an in silico analysis of the Zophobas atratus transcriptome to identify putative antimicrobial peptides and identified several antimicrobial peptide candidates. Among them, we assessed the antimicrobial and anti-inflammatory activities of zophobacin 1 that was selected bio-informatically based on its physicochemical properties against microorganisms and mouse macrophage Raw264.7 cells. Zophobacin 1 showed antimicrobial activities against microorganisms without inducing hemolysis and decreased the nitric oxide production of the lipopolysaccharide-induced Raw264.7 cells. Moreover, ELISA and Western blot analysis revealed that zophobacin 1 reduced expression levels of pro-inflammatory enzymes such as inducible nitric oxide synthase (iNOS) and cyclooxygenase-2 (COX-2). We also investigated expression of pro-inflammatory cytokines (interleukin-6 and interleukin-1β) production through quantitative real time-PCR and ELISA. Zophobacin 1 markedly reduced the expression level of cytokines through the regulation of mitogen-activated protein kinases (MAPKs) and nuclear factor kappa B (NF-κB) signaling. We confirmed that zophobacin 1 bound to bacterial cell membranes via a specific interaction with lipopolysaccharides. These data suggest that zophobacin 1 could be promising molecules for development as antimicrobial and anti-inflammatory therapeutic agents.

• Science of Emotion and Sensibility
• /
• v.13 no.1
• /
• pp.47-60
• /
• 2010

Most of the researches about classification usually have used kNN(k-Nearest Neighbor), SVM(Support Vector Machine), which are known as learn-based model, and Bayesian classifier, NNA(Neural Network Algorithm), which are known as statistics-based methods. However, there are some limitations of space and time when classifying so many web pages in recent internet. Moreover, most studies of classification are using uni-gram feature representation which is not good to represent real meaning of words. In case of Korean web page classification, there are some problems because of korean words property that the words have multiple meanings(polysemy). For these reasons, LSA(Latent Semantic Analysis) is proposed to classify well in these environment(large data set and words' polysemy). LSA uses SVD(Singular Value Decomposition) which decomposes the original term-document matrix to three different matrices and reduces their dimension. From this SVD's work, it is possible to create new low-level semantic space for representing vectors, which can make classification efficient and analyze latent meaning of words or document(or web pages). Although LSA is good at classification, it has some drawbacks in classification. As SVD reduces dimensions of matrix and creates new semantic space, it doesn't consider which dimensions discriminate vectors well but it does consider which dimensions represent vectors well. It is a reason why LSA doesn't improve performance of classification as expectation. In this paper, we propose new LSA which selects optimal dimensions to discriminate and represent vectors well as minimizing drawbacks and improving performance. This method that we propose shows better and more stable performance than other LSAs' in low-dimension space. In addition, we derive more improvement in classification as creating and selecting features by reducing stopwords and weighting specific values to them statistically.

• Journal of Plant Biotechnology
• /
• v.43 no.3
• /
• pp.359-366
• /
• 2016

Tradescantia is a perennial plant in the family of Commelinaceae. It is known to be sensitive to radiation. In this study, Tradescantia BNL 4430 was irradiated with gamma radiation at doses of 50 to 1,000 mGy in a phytotron equipped with a $^{60}Co$ radiation source at Korea Atomic Energy Research Institute, Korea. At 13 days after irradiation, we extracted RNA from irradiated floral tissues for RNA-seq. Transcriptome assembly produced a total of 77, 326 unique transcripts. In plantlets exposed to 50, 250, 500, and 1000 mGy, the numbers of up-regulated genes with more than 2-fold of expression compared that in the control were 116, 222, 246, and 308, respectively. Most of the up-regulated genes induced by 50 mGy were heat shock proteins (HSPs) such as HSP 70, indicating that protein misfolding, aggregation, and translocation might have occurred during radiation stress. Similarly, highly up-regulated transcripts of the IQ-domain 6 were induced by 250 mGy, KAR-UP oxidoreductase 1 was induced by 500 mGy, and zinc transporter 1 precursor was induced by 1000 mGy. Reverse transcriptase (RT) PCR and quantitative real time PCR (qRT-PCR) further validated the increased mRNA expression levels of selected genes, consistent with DEG analysis results. However, 2.3 to 97- fold higher expression activities were induced by different doses of radiation based on qRT-PCR results. Results on the transcriptome of Tradescantia in response to radiation might provide unique identifiers to develop in situ monitoring kit for measuring radiation exposure around radiation facilities.

• Journal of Life Science
• /
• v.23 no.1
• /
• pp.116-124
• /
• 2013

The aim was to examine resistance exercise-related genes after 8 weeks of resistance training. Thirty-two male Sprague-Dawley rats were divided into four groups: 4 weeks sedentary (4 wks CON, n=8), 8 weeks sedentary (8 wks CON, n=8), 4 weeks exercise training (4 wks REG, n=8), and 8 weeks exercise training (8 wks REG, n=8). The rats were trained to climb a 1-m vertical incline (85-degree), with weights secured to their tails. They climbed 10 times, 3 days per week, for 8 consecutive weeks. Skeletal muscle was taken from the flexor halucis longus after the exercise training. After separating the total RNA, large-scale gene expression was investigated by beadarray (Illumina RatRef-12 Expression BeadChip) analysis, and qPCR was used to inspect the beadarray data and to analyze the RNA quantitatively. The detection p-value for the genes was p<0.01, the M-value {M=$log_2$(condition)-$log_2$(reference)} was >1.0, and the DiffScore was >20. In total, the expression of 30 genes significantly increased 4 weeks after the exercise training, and the expression of six genes decreased. At 8 weeks, the expression of five genes significantly increased and that of 12 decreased. Several genes are potentially involved in resistance exercise and muscle hypertrophy, including 1) regulation of cell growth (IGFBP1, PLA2G2A, OKL38); 2) myogenesis (CSRP3); 3) tissue regeneration and muscle development (MUSTN1, MYBPH); 4) hypertrophy (CYR61, ATF3, NR4A3); and 5) glucose metabolism (G6PC, PCK1). These results may help to explain previously reported physiological changes of the skeletal muscle and suggest new avenues for further investigation.