Search

Abstract The application of deep learning techniques for the analysis of neuroimaging has been increasing recently. The 3D Convolutional Neural Network (CNN) technology, which is commonly adopted to encode volumetric information, requires a large number of datasets. However, due to the nature of the medical domain, there are limitations in the number of data available. This is because the cost of acquiring imaging is expensive and the use of personnel to annotate diagnostic labels is resource-intensive. For these reasons, several prior studies have opted to use comparatively lighter 2D CNNs instead of the complex 3D CNN technology. They analyze using projected 2D datasets created from representative slices extracted from 3D volumetric imaging. However, this approach, by selecting only projected 2D slices from the entire volume, reflects only partial volumetric information. This poses a risk of developing lesion diagnosis systems without a deep understanding of the interrelations among volumetric data. We propose a novel 3D-to-2D knowledge distillation framework that utilizes not only the projected 2D dataset but also the original 3D volumetric imaging dataset. This framework is designed to employ volumetric prior knowledge in training 2D CNNs. Our proposed method includes three modules: (i) a 3D teacher network that encodes volumetric prior knowledge from the 3D dataset, (ii) a 2D student network that encodes partial volumetric information from the 2D dataset, and aims to develop an understanding of the original volumetric imaging, and (iii) a distillation loss introduced to reduce the gap in the graph representation expressing the relationship between data in the feature embedding spaces of (i) and (ii), thereby enhancing the final performance. The effectiveness of our proposed method is demonstrated by improving the classification performance orthogonally across various 2D projection methods on two datasets from the 123I-DaTscan SPECT and 18 F-AV133 PET from Parkinson’s Progression Markers Initiative (PPMI). Notably, when our approach is applied to the FuseMe approach, it achieves an F1 score of 98.30%, which is higher than that of the 3D teacher network (97.66%).

Five new species of the genus Phaenocarpa Foester, 1863 (Braconidae: Alysiinae) are described and illustrated: P. acutidentata Sohn & van Achterberg, sp. nov., P. tacitoides Sohn & van Achterberg, sp. nov., P. setosa Sohn & van Achterberg, sp. nov., P. tanycauda Sohn & van Achterberg, sp. nov., and P. angusticeps Sohn & van Achterberg, sp. nov. Additionally, P. tacita Stelfox, 1941 is recorded for the first time from South Korea. The barcode region of mitochondrial cytochrome c oxidase I (COI) was also analyzed for the seven congeneric species including one from GenBank. In addition, an identification key for the Phaenocarpa species recorded in Korea is provided.

Abstract Background Immune checkpoint therapy (ICT) provides durable responses in select cancer patients, yet resistance remains a significant challenge, prompting the exploration of underlying molecular mechanisms. Tyrosylprotein sulfotransferase-2 (TPST2), known for its role in protein tyrosine O-sulfation, has been suggested to modulate the extracellular protein-protein interactions, but its specific role in cancer immunity remains largely unexplored. Methods To explore tumor cell-intrinsic factors influencing anti-PD1 responsiveness, we conducted a pooled loss-of-function genetic screen in humanized mice engrafted with human immune cells. The responsiveness of cancer cells to interferon-γ (IFNγ) was estimated by evaluating IFNγ-mediated induction of target genes, STAT1 phosphorylation, HLA expression, and cell growth suppression. The sulfotyrosine-modified target gene of TPST2 was identified by co-immunoprecipitation and mass spectrometry. The in vivo effects of TPST2 inhibition were evaluated using mouse syngeneic tumor models and corroborated by bulk and single-cell RNA sequencing analyses. Results Through in vivo genome-wide CRISPR screening, TPST2 loss-of-function emerged as a potential enhancer of anti-PD1 treatment efficacy. TPST2 suppressed IFNγ signaling by sulfating IFNγ receptor 1 at Y397 residue, while its downregulation boosted IFNγ-mediated signaling and antigen presentation. Depletion of TPST2 in cancer cells augmented anti-PD1 antibody efficacy in syngeneic mouse tumor models by enhancing tumor-infiltrating lymphocytes. RNA sequencing data revealed TPST2’s inverse correlation with antigen presentation, and increased TPST2 expression is associated with poor prognosis and altered cancer immunity across cancer types. Conclusions We propose TPST2’s novel role as a suppressor of cancer immunity and advocate for its consideration as a therapeutic target in ICT-based treatments.

The genus Monoctonus Haliday, 1833 is a small group which consists of 24 species worldwide. In South Korea, Chang and Youn (1983) recorded one species, M. similis Starý & Schlinger, 1967, but the evidence for identification of this species is doubtful and further confirmation is required (personal communication with Prof. Jong-Cheol Paik).An additional Monoctonus species is recorded as new to science from South Korea. Descriptions and illustrations of the new species –Monoctonus koreanus sp. nov. – are provided, together with its mitochondrial cytochrome c oxidase subunit I (COI) data and phylogenetic position. A key to the female of the two species present in Korea is provided.

The genus Trioxys Haliday, 1833 consists of more than 80 species worldwide with three species being recorded in South Korea. In this study, we report the first observation of the two additional species, T. liui Chou & Chou, 1993 from Takecallis arundinariae (Essig, 1917) on Phyllostachys bambusoides Siebold & Zucc., 1843 and T. remaudierei Starý & Rakhshani, 2017 from T. taiwana (Takahashi, 1926) on Sasa borealis (Hack.) Makino & Shibata, 1901.Trioxys liui and T. remaudierei are described and reported with phototographs of the diagnostic morphological characters and the mitochondrial cytochrome c oxidase subunit I (COI) data (barcode region) and Bayesian tree of the phylogenetic analysis amongst the closely-related taxa are provided.

Osteoarthritis (OA) is characterized by irreversible joint destruction, pain, and dysfunction. Piper longum L. [Piperaceae] (PL) is an East Asian herbal medicine with reported anti-inflammatory, analgesic, antioxidant, anti-stress, and anti-osteoporotic effects. This study aimed to evaluate the efficacy of PL in inhibiting pain and progressive joint destruction in OA based on its anti-inflammatory activity, and to explore its potential mechanisms using in vivo and in vitro models of OA. We predicted the potential hub targets and signaling pathways of PL through network analysis and molecular docking. Network analysis results showed that the possible hub targets of PL against OA were F2R, F3, MMP1, MMP2, MMP9, and PTGS2. The molecular docking results predicted strong binding affinities for the core compounds in PL: piperlongumine, piperlonguminine, and piperine. In vitro experiments showed that PL inhibited the expression of LPS-induced pro-inflammatory factors, such as F2R, F3, IL-1β, IL-6, IL-17A, MMP-1, MMP-2, MMP-3, MMP-9, MMP-13, NOS2, PTGS2, PGE2, and TNF-β. These mechanisms and effects were dose-dependent in vivo models. Furthermore, PL inhibited cartilage degradation in an OA-induced rat model. Thus, this study demonstrated that multiple components of PL may inhibit the multilayered pathology of OA by acting on multiple targets and pathways. These findings highlight the potential of PL as a disease-modifying OA drug candidate, which warrants further investigation.

The genus Toxares Haliday, 1840 is a small taxon of Aphidiinae, consisting four valid species in the world. One Toxares species is recorded as new to science from South Korea, in this study. Descriptions and illustrations of the new species, T. koreanus sp. nov., are provided, together with their mitochondrial cytochrome c oxidase subunit I (COI) and D2 region of the nuclear gene for 28S rRNA (28S) sequences. The phylogenetic tree reconstructed using a combination of COI and 28S revealed the phylogenetic position of the genus Toxares within Aphidiinae.

Separatatus megagnathus sp. nov. is recorded as new to science from South Korea. Due to this record, the genus Separatatus Chen & Wu, 1994 (Braconidae: Alysiinae) is recognized for the first time from South Korea. The genus and species are described and illustrated herein plus an identification key including the Korean new species is provided. In addition, the DNA barcode region of the mitochondrial cytochrome c oxidase subunit I (COI) has been analyzed for the new species.

The stock index is actively used for the realization of profits using derivatives and via the hedging of assets; hence, the prediction of the index is important for market participants. As market uncertainty has increased during the COVID-19 pandemic and with the rapid development of data engineering, a situation has arisen wherein extensive amounts of information must be processed at finer time intervals. Addressing the prevalent issues of difficulty in handling multivariate high-frequency time-series data owing to multicollinearity, resource problems in computing hardware, and the gradient vanishing problem due to the layer stacking in recurrent neural network (RNN) series, a novel algorithm is developed in this study. For financial market index prediction with these highly complex data, the algorithm combines ResNet and a variable-wise attention mechanism. To verify the superior performance of the proposed model, RNN, long short-term memory, and ResNet18 models were designed and compared with and without the attention mechanism. As per the results, the proposed model demonstrated a suitable synergistic effect with the time-series data and excellent classification performance, in addition to overcoming the data structure constraints that the other models exhibit. Having successfully presented multivariate high-frequency time-series data analysis, this study enables effective investment decision making based on the market signals.

Several species of the genus Aphidius are well known as commercial biocontrol agents of pest aphids, and more than 130 species of the genus have been recorded worldwide. To date, only 15 Aphidius species have been recorded in South Korea. Using the DNA barcode region (ca. 658 bp) of mitochondrial cytochrome c oxidase subunit I (COI), we amplified sequences of 15 Aphidius Korean species, aligned them in combination with 23 reference species retrieved from GenBank for comparison and identification, and then reconstructed a barcode phylogeny by the neighbour-joining method. As a result, three Aphidius species were found to be new to Science. Descriptions and illustrations of the three species new to Science – Aphidius longicarpus sp. nov., A. longistigmus sp. nov., and A. asiaticus sp. nov. – are provided, together with their phylogenetic position within the genus Aphidius. In addition, a redescription of A. areolatus, a parasitoid of maple aphids (Peryphillus spp.), is also given.

Background : Tacrolimus (TAC) is a main therapy for liver transplantation (LT) patients, but it has side effects such as chronic nephrotoxicity that progressively aggravate renal function. The purpose of this study was to retrospectively compare the renal function between a TAC group and a combination of everolimus and reduced TAC (EVR-TAC) group after deceased donor liver transplantation (DDLT). Methods : The study comprised 131 patients who underwent DDLT between January 2013 and April 2018 at our institution. They received TAC or EVR-TAC after DDLT. Everolimus (EVR) was introduced between 1 and 6 months after DDLT. Results : Thirty-six of 131 patients (27.5%) received EVR-TAC. The incidence of chronic kidney disease (CKD; estimated glomerular filtration rate,

Hybrid broadcasting utilizing heterogeneous networks can provide not only Audio/Video(A/V)-based unidirectional broadcast services but also various interactive A/V and data services via both broadcast and communication networks. Customers of hybrid broadcasting services can then consume A/V content that is transmitted from broadcasting stations and data services relevant to A/V content through communication networks, such as the Internet. Since interactive data can be widely ingested in various countries and regions, it is advantageous for it to be customized for personalized services depending on consumer preferences and locations. However, the current broadcasting standard has not yet defined a standardized method of inserting content into a broadcasting program. Thus, the research on content insertion technology using heterogeneous networks is actively progressing. The most important technical challenge of inserting content is rendering a smooth replacement of on-air broadcast content with media data through communication networks and an accurate return to the over-the-air broadcast content consistent with the rendering timeline using legacy receivers. Since most legacy receivers have one decoder, both the inserted content and the broadcast content have to be handled by one decoder. When two video streams are processed by one decoder, decoding and rendering problems due to the buffer overflow may occur, and buffer management of the decoder is required to solve this problem. Therefore, this paper proposes a content insertion method of using a frame control-based receiver buffer model to enable smooth and accurate rendering of inserted and on-air content without any blackout scenes.

Dementia is a degenerative disease that is increasingly prevalent in an aging society. Alzheimer’s disease (AD), the most common type of dementia, is best mitigated via early detection and management. Deep learning is an artificial intelligence technique that has been used to diagnose and predict diseases by extracting meaningful features from medical images. The convolutional neural network (CNN) is a representative application of deep learning, serving as a powerful tool for the diagnosis of AD. Recently, vision transformers (ViT) have yielded classification performance exceeding that of CNN in some diagnostic image classifications. Because the brain is a very complex network with interrelated regions, ViT, which captures direct relationships between images, may be more effective for brain image analysis than CNN. Therefore, we propose a method for classifying dementia images by applying 18F-Florbetaben positron emission tomography (PET) images to ViT. Data were evaluated via binary (normal control and abnormal) and ternary (healthy control, mild cognitive impairment, and AD) classification. In a performance comparison with the CNN, VGG19 was selected as the comparison model. Consequently, ViT yielded more effective performance than VGG19 in binary classification. However, in ternary classification, the performance of ViT cannot be considered excellent. These results show that it is hard to argue that the ViT model is better at AD classification than the CNN model.

Infection with Mycobacterium leprae, the causative organism of leprosy, is still endemic in numerous parts of the world including the southwestern United States. The broad variation of symptoms in the leprosy disease spectrum range from the milder tuberculoid leprosy (paucibacillary) to the more severe and disfiguring lepromatous leprosy (multibacillary). The established thinking in the health community is that host response, rather than M. leprae strain variation, is the reason for the range of disease severity. More recent discoveries suggest that macrophage polarization also plays a significant role in the spectrum of leprosy disease but to what degree it contributes is not fully established. In this study, we aimed to analyze if different strains of M. leprae elicit different transcription responses in human macrophages, and to examine the role of macrophage polarization in these responses.Genomic DNA from three different strains of M. leprae DNA (Strains NHDP, Br4923, and Thai-53) were used to stimulate human macrophages under three polarization conditions (M1, M1-activated, and M2).Transcriptome analysis revealed a large number of differentially expressed (DE) genes upon stimulation with DNA from M. leprae strain Thai-53 compared to strains NHDP and Br4923, independent of the macrophage polarization condition. We also found that macrophage polarization affects the responses to M. leprae DNA, with up-regulation of numerous interferon stimulated genes.These findings provide a deeper understanding of the role of macrophage polarization in the recognition of M. leprae DNA, with the potential to improve leprosy treatment strategies.

Background : Patients who undergo reoperation after living-donor liver transplantation (LDLT) have poor outcomes. However, the specific outcomes of patients undergoing reoperation due to gastrointestinal (GI) tract-related complications following adult LDLT are relatively unknown. In the present study, we investigated the relationship between the causes and outcomes of reoperation after LDLT and classified the risk groups. Methods : We performed a retrospective analysis of 506 patients who underwent LDLT at Samsung Medical Center in Seoul between 2010 and 2016.Results : Among 506 adult LDLT recipients, 98 (19.4%) underwent reoperation. The causes for reoperation included bleeding (n=39, 39.8%), vascular complications (n=26, 26.5%), wound complications (n=12, 12.2%), bile leakage (n=7, 7.1%), GI tract complications (n=6, 6.1%), and others (n=8, 8.1%). Based on a multivariate analysis, we identified prolonged operation time, hospitalization days, and a history of previous hepatocellular carcinoma-related operation as independent risk factors for reoperation. Patient survival after 3 months, 1 year, 3 years, and 5 years was 96.3%, 90.6%, 82.5%, and 79.4% in the non-reoperation group and 95.9%, 82.7%, 72.8% and 69.3% in the reoperation group, respectively. Patient survival in the reoperation group was significantly lower than that in the non-reoperation group (P=0.018). In the reoperation group, the survival rates of patients with GI tract-related complications-including bile leakage and GI tract complications-were significantly worse than those of patients with non-GI tract-related complications such as bleeding, vascular complications, and wound complications (P

Cryptocurrencies are highly volatile investment assets and are difficult to predict. In this study, various cryptocurrency data are used as features to predict the log-return price of major cryptocurrencies. The original contribution of this study is the selection of the most influential major features for each cryptocurrency using the volatility features of cryptocurrency, derived from the autoregressive conditional heteroskedasticity (ARCH) and generalized autoregressive conditional heteroskedasticity (GARCH) models, along with the closing price of the cryptocurrency. In addition, we sought to predict the log-return price of cryptocurrencies by implementing various types of time-series model. Based on the selected major features, the log-return price of cryptocurrency was predicted through the autoregressive integrated moving average (ARIMA) time-series prediction model and the artificial neural network-based time-series prediction model. As a result of log-return price prediction, the neural-network-based time-series prediction models showed superior predictive power compared to the traditional time-series prediction model.

BackgroundsInadequate liver volume and weight is a major source of morbidity and mortality after adult living donor liver transplantation (LDLT). The purpose of our study was to investigate HCC recurrence, graft failure, and patient survival according to change in right liver graft weight after histidine-tryptophan-ketoglutarate (HTK) solution perfusion in LDLT.MethodsTwo hundred twenty-eight patients underwent LDLT between 2013 and 2017. We calculated the change in graft weight by subtracting pre-perfusion graft weight from post-perfusion graft weight. Patients with increased graft weight were defined as the positive group, and patients with decreased graft weight were defined as the negative group.ResultsAfter excluding patients who did not meet study criteria, 148 patients underwent right or extended right hepatectomy. The negative group included 89 patients (60.1%), and the positive group included 59 patients (39.9%). Median graft weight change was -28 g (range; -132–0 g) in the negative group and 21 g (range; 1–63 g) in the positive group (P

18F-florbetaben (FBB) positron emission tomography is a representative imaging test that observes amyloid deposition in the brain. Compared to delay-phase FBB (dFBB), early-phase FBB shows patterns related to glucose metabolism in 18F-fluorodeoxyglucose perfusion images. The purpose of this study is to prove that classification accuracy is higher when using dual-phase FBB (dual FBB) versus dFBB quantitative analysis by using machine learning and to find an optimal machine learning model suitable for dual FBB quantitative analysis data. The key features of our method are (1) a feature ranking method for each phase of FBB with a cross-validated F1 score and (2) a quantitative diagnostic model based on machine learning methods. We compared four classification models: support vector machine, naïve Bayes, logistic regression, and random forest (RF). In composite standardized uptake value ratio, RF achieved the best performance (F1: 78.06%) with dual FBB, which was 4.83% higher than the result with dFBB. In conclusion, regardless of the two quantitative analysis methods, using the dual FBB has a higher classification accuracy than using the dFBB. The RF model is the machine learning model that best classifies a dual FBB. The regions that have the greatest influence on the classification of dual FBB are the frontal and temporal lobes.

This paper introduces methodologies in forecasting oil prices (Brent and WTI) with multivariate time series of major S&P 500 stock prices using Gaussian process modeling, deep learning, and vine copula regression. We also apply Bayesian variable selection and nonlinear principal component analysis (NLPCA) for data dimension reduction. With a reduced number of important covariates, we also forecast oil prices (Brent and WTI) with multivariate time series of major S&P 500 stock prices using Gaussian process modeling, deep learning, and vine copula regression. To apply real data to the proposed methods, we select monthly log returns of 2 oil prices and 74 large-cap, major S&P 500 stock prices across the period of February 2001–October 2019. We conclude that vine copula regression with NLPCA is superior overall to other proposed methods in terms of the measures of prediction errors.

Objectives We investigated the endoplasmic reticulum (ER) stress markers C/EBP homologous protein (CHOP) and glucose-regulated protein (GRP) 78, as well as the inflammatory factors nuclear factor (NF)-κB and IκBα, to assess how social defeat stress induces myocardial injury. Furthermore, we evaluated the protective effects of the ER stress inhibitor 4-phenylbutyric acid (PBA) on myocardial injury in mice. Methods Adult mice were divided into control, control + PBA, social defeat, and social defeat + PBA groups. The social defeat and social defeat + PBA groups were exposed to social defeat stress for 10 days. Cardiac tissues from all groups were analyzed after social defeat stress. H 9 C 2 cells were used to detect the role of the ER stress agonist thapsigargin on expression of ER stress and inflammatory markers. Results Social defeat stress promoted apoptosis of cardiomyocytes, increased CHOP, NF-κB and, phospho-NF-κB protein expression, and decreased GRP78 and IκBα protein expression. Moreover, PBA significantly reversed these changes and attenuated thapsigargin-induced increased expression of CHOP and phospho-NF-κB, and decreased IκBα expression in H 9 C 2 cells. Conclusions Social defeat stress initiates ER stress, promotes expression of inflammatory factors, and induces myocardial injury. Inhibiting ER stress could protect the myocardium from social defeat stress-induced myocardial injury.

To infer the introduction sources of the oriental fruit fly, Bactrocera dorsalis, we used a mitochondrial marker to reconstruct the haplotype network and 15 microsatellite loci to reveal genetic structure and relationships between the geographically or temporally different collections from Asia. We performed Approximate Bayesian computations to infer a global origin and a source of the quarantine collections found in Korea. As a result, the 40 populations were divided into three groups, of which genetic similarity is not related to the geographic vicinity. Korean samples had a similar genetic structure to Taiwan and Thailand ones. Our results suggest that the place of origin of the B. dorsalis specimens found in Korea’s border quarantine is likely to be Taiwan or Thailand. As the global origin of B. dorsalis, we estimated that Taiwan and Thailand were most likely the global origins of Southeast Asian populations by testing hypothetical scenarios by the approximate Bayesian computation analyses. Our results will allow easier identification of the source region of the forthcoming invasion of quarantined B. dorsalis specimens.

Along with the fourth industrial revolution, research in the biomedical engineering field is being actively conducted. Among these research fields, the brain–computer interface (BCI) research, which studies the direct interaction between the brain and external devices, is in the spotlight. However, in the case of electroencephalograph (EEG) data measured through BCI, there are a huge number of features, which can lead to many difficulties in analysis because of complex relationships between features. For this reason, research on BCIs using EEG data is often insufficient. Therefore, in this study, we develop the methodology for selecting features for a specific type of BCI that predicts whether a person correctly detects facial expression changes or not by classifying EEG-based features. We also investigate whether specific EEG features affect expression change detection. Various feature selection methods were used to check the influence of each feature on expression change detection, and the best combination was selected using several machine learning classification techniques. As a best result of the classification accuracy, 71% of accuracy was obtained with XGBoost using 52 features. EEG topography was confirmed using the selected major features, showing that the detection of changes in facial expression largely engages brain activity in the frontal regions.

This paper examines the relationship of the leading financial assets, Bitcoin, Gold, and S&P 500 with GARCH-Dynamic Conditional Correlation (DCC), Nonlinear Asymmetric GARCH DCC (NA-DCC), Gaussian copula-based GARCH-DCC (GC-DCC), and Gaussian copula-based Nonlinear Asymmetric-DCC (GCNA-DCC). Under the high volatility financial situation such as the COVID-19 pandemic occurrence, there exist a computation difficulty to use the traditional DCC method to the selected cryptocurrencies. To solve this limitation, GC-DCC and GCNA-DCC are applied to investigate the time-varying relationship among Bitcoin, Gold, and S&P 500. In terms of log-likelihood, we show that GC-DCC and GCNA-DCC are better models than DCC and NA-DCC to show relationship of Bitcoin with Gold and S&P 500. We also consider the relationships among time-varying conditional correlation with Bitcoin volatility, and S&P 500 volatility by a Gaussian Copula Marginal Regression (GCMR) model. The empirical findings show that S&P 500 and Gold price are statistically significant to Bitcoin in terms of log-return and volatility.

In high-dimensional data, the performances of various classifiers are largely dependent on the selection of important features. Most of the individual classifiers with the existing feature selection (FS) methods do not perform well for highly correlated data. Obtaining important features using the FS method and selecting the best performing classifier is a challenging task in high throughput data. In this article, we propose a combination of resampling-based least absolute shrinkage and selection operator (LASSO) feature selection (RLFS) and ensembles of regularized regression (ERRM) capable of dealing data with the high correlation structures. The ERRM boosts the prediction accuracy with the top-ranked features obtained from RLFS. The RLFS utilizes the lasso penalty with sure independence screening (SIS) condition to select the top k ranked features. The ERRM includes five individual penalty based classifiers: LASSO, adaptive LASSO (ALASSO), elastic net (ENET), smoothly clipped absolute deviations (SCAD), and minimax concave penalty (MCP). It was built on the idea of bagging and rank aggregation. Upon performing simulation studies and applying to smokers’ cancer gene expression data, we demonstrated that the proposed combination of ERRM with RLFS achieved superior performance of accuracy and geometric mean.

In high-dimensional gene expression data analysis, the accuracy and reliability of cancer classification and selection of important genes play a very crucial role. To identify these important genes and predict future outcomes (tumor vs. non-tumor), various methods have been proposed in the literature. But only few of them take into account correlation patterns and grouping effects among the genes. In this article, we propose a rank-based modification of the popular penalized logistic regression procedure based on a combination of ℓ 1 and ℓ 2 penalties capable of handling possible correlation among genes in different groups. While the ℓ 1 penalty maintains sparsity, the ℓ 2 penalty induces smoothness based on the information from the Laplacian matrix, which represents the correlation pattern among genes. We combined logistic regression with the BH-FDR (Benjamini and Hochberg false discovery rate) screening procedure and a newly developed rank-based selection method to come up with an optimal model retaining the important genes. Through simulation studies and real-world application to high-dimensional colon cancer gene expression data, we demonstrated that the proposed rank-based method outperforms such currently popular methods as lasso, adaptive lasso and elastic net when applied both to gene selection and classification.

Over the last decade, high dimensional data have been popularly paid attention to in bioinformatics. These data increase the likelihood of detecting the most promising novel information. However, there are limitations of high-performance computing and overfitting issues. To overcome the issues, alternative strategies need to be explored for the detection of true important features. A two-stage approach, filtering and variable selection steps, has been receiving attention. Filtering methods are divided into two categories of individual ranking and feature subset selection methods. Both have issues with the lack of consideration for joint correlation among features and computing time of an NP-hard problem. Therefore, we proposed a new filter ranking method (PF) using the elastic net penalty with sure independence screening (SIS) based on resampling technique to overcome these issues. We demonstrated that SIS-LASSO, SIS-MCP, and SIS-SCAD with the proposed filtering method achieved superior performance of not only accuracy, AUROC, and geometric mean but also true positive detection compared to those with the marginal maximum likelihood ranking method (MMLR) through extensive simulation studies. In addition, we applied it in a real application of colon and lung cancer gene expression data to investigate the classification performance and power of detecting true genes associated with colon and lung cancer.

We have maintained the customer grade system that is being implemented to customers with excellent performance through customer segmentation for years. Currently, financial institutions that operate the customer grade system provide similar services based on the score calculation criteria, but the score calculation criteria vary from the financial institution to financial institution. In this study, we create a machine learning prediction model using items and added items that are based on the current customer grade of our bank,- and the purpose is an optimal model that considers the adequacy of existing variables and the validity of additional variables through comparison between models. Using Lasso, Elastic net and Multinomial Logistic Regression, Decision Tree, Random Forest, and Support Vector Machine, we propose that the best model be found and gradually applied to customer grade calculation criteria.

The default risk, one of the main risk factors for bonds, should be measured and reflected in the bond yield. Particularly, in the case of financial companies that treat bonds as a major product, failure to properly identify and filter customers' workout status adversely affects returns. This study proposes a two-stage classification algorithm for workout prediction based on the history data of individual customers such as transaction details of financial companies secured after loans, which is collected over 10 years. The first stage is to rank variables that are closely related to the workout application based on feature selection. In the second step, the first to nth cumulative variables input to each machine learning method generate n candidate classifiers, respectively. Among the total candidates, the model with the highest classification accuracy was selected as the optimal one, which is the Gradient Boost combined with F-score-based feature selection.

Corrosion is a factor that decreases the ultimate strength of a structure, which can lead to structural failure in an aged structure. This study proposes a reliable method to predict the ultimate strength reduction caused by pitting corrosion of a structure at various ages. Hence, the objectives of this study are as follows: (1) finding relationship between the ultimate strength reduction and pit depth and (2) finding relationship between the ultimate strength reduction and the age of structure. The presented method has six steps and was assessed by applying it to the simplified example of a thin-plate panel structure. The ultimate strengths of the panel with different pit depths were calculated using finite element analysis. The effects of material and geometric nonlinearities as well as the initial imperfections and buckling were also considered. A linear regression model, representing the relationship between the ultimate strength reduction and pit depth, was then developed. Furthermore, using the experimental data, the weighted mean value of ultimate strength reduction was acquired for different age groups of the panel. The introduced stepwise approach contemplates the nonlinearity of pitting corrosion with time by considering the experimental data. As corroborated by the successful application of the presented methodology, this method is more accurate, compared to the accepted practices available in the literature.

With advent of various mobile devices with powerful networking and computing capabilities, the users' demand to enjoy live video streaming services such as IPTV with mobile devices has been increasing rapidly. However, it is challenging to get over the degradation of service quality due to data loss caused by the handover. Although many handover schemes were proposed at protocol layers below the application layer, they inherently suffer from data loss while the network is being disconnected during the handover. We therefore propose an efficient application-layer handover scheme to support seamless mobility for P2P live streaming. By simulation experiments, we show that the P2P live streaming system with our proposed handover scheme can improve the playback continuity significantly compared to that without our scheme.