Search

With the increase in the aging population of many countries, the prevalence of neovascular age-related macular degeneration (nAMD) is expected to increase. Morphological parameters such as intraretinal fluid (IRF), subretinal fluid (SRF), subretinal hyperreflective material (SHRM), and pigment epithelium detachment (PED) of spectral-domain optical coherence tomography (SD-OCT) images are vital markers for proper treatment of nAMD, especially to get the information of treatment response to determine the proper treatment interval and switching of anti-vascular endothelial growth factor (VEGF) agents. For the precise evaluation of the change in nAMD lesions and patient-specific treatment, quantitative evaluation of the lesions in the OCT volume scans is necessary. However, manual segmentation requires many resources, and the number of studies of automatic segmentation is increasing rapidly. Improving automated segmentation performance in SD-OCT visual results requires long-range contextual inference of spatial information between retinal lesions and layers. This paper proposes a GAGUNet (graph convolution network (GCN)-assisted attention-guided UNet) model with a novel global reasoning module considering these points. The dataset used in the main experiment of this study underwent rigorous review by a retinal specialist from Konkuk University Hospital in Korea, contributing to both data preprocessing and validation to ensure a qualitative assessment. We conducted experiments on the RETOUCH dataset as well to demonstrate the scalability of the proposed model. Overall, our model demonstrates superior performance over the baseline models in both quantitative and qualitative evaluations. • We demonstrate the limitations of the baseline in retinal segmentation. • We propose a novel model for automated retinal OCT segmentation. • The proposed model performs higher than the baseline in retinal segmentation. [ABSTRACT FROM AUTHOR]

Members of the armed forces rely on having an effective and efficient medical evacuation (MEDEVAC) process for evacuating casualties from the battlefield to medical treatment facilities during combat operations. Determining which MEDEVAC units to task to respond to incoming requests for service, known as the MEDEVAC dispatching problem, is a natural sequential decision making problem and is the focus of this paper. Unlike previous research in this area, this paper incorporates triage classification errors and the possibility of having blood transfusion kits on board select MEDEVAC units. A discounted, infinite-horizon continuous-time Markov decision process model is formulated to examine the MEDEVAC dispatching problem and to compare generated dispatching policies to the currently practiced United States military policy. We utilize an approximate dynamic programming (ADP) technique that leverages a random forest value function approximation within an approximate policy iteration algorithmic framework to develop high-quality policies for both small-scale and large-scale problem instances. A representative planning scenario involving joint combat operations in South Korea is developed and utilized to investigate the differences between the various policies. Results from the analysis indicate that applying ADP techniques can improve current practices by as much as 29% with regard to a life-saving performance metric. This research is of particular interest to the military medical community and can inform the procedures of future military MEDEVAC operations. [ABSTRACT FROM AUTHOR]

• Employs non-identical parallel machines in manufacturing with multiple energy sources. • Designs a bi-objective model for a high-mix, low-volume production scheduling problem. • Provides Pareto-optimal solutions that minimize both total cost and carbon emissions. • Validates a model with real-world application to a manufacturer in Ulsan, South Korea. • Analyzes the effects of price fluctuations, different combinations of energy sources. This paper addresses the production scheduling problem with non-identical parallel machines in a high-mix, low-volume production environment with due dates, considering carbon emissions savings and diverse types of energy sources for machine operation. To this end, we present a bi-objective mixed-integer programming model that minimizes both the total production-related cost and the total carbon emissions in the production process to determine the optimal production strategy. For each machine, we consider various types of energy sources with different electricity generation costs and different amounts of carbon emissions. The proposed model is validated with an application to a manufacturer in Ulsan, the industrial capital of the Republic of Korea. Our results showcase the potential use of non-identical parallel machines to minimize the total cost and carbon emissions in green manufacturing. Furthermore, the results identify the trade-off between different energy sources and the total cost under different carbon emissions limits. Generally, as the carbon emissions limit increases, our proposed model tends to replace natural gas with coal to minimize the total cost. Also, we perform sensitivity analyses with respect to the energy price of different types of energy sources combined with nuclear and renewable energy sources. We find that coal provides more stability than natural gas in terms of the total cost, particularly when there are fluctuations in the price for coal and natural gas. Additionally, we determine the optimal combination of energy sources for various carbon emissions limits, aiming to minimize the total cost while simultaneously satisfying all production-related and environmental constraints. [ABSTRACT FROM AUTHOR]

• Developed a Neural Network foreign exchange trading system using CCS-IRS basis. • The NN foreign exchange trading model outperforms XGBoost and LR model in USDKRW. • The NN trading model outperforms a buy-n-hold and sell-n-hold strategies in USDKRW. • The CCS-IRS basis replaces the CDS premium as a new leading indicator in USDKRW. This study aims to develop a foreign exchange trading model with a new leading indicator based on cross-currency swap (CCS) and interest rate swap (IRS) rates. We derive the correlation dimension from the chaos analysis, use Neural Networks (NNs) to optimize foreign exchange trading, and then compare the performance of the NNs model with XGBoost and Logistic Regression model (LR). The CCS-IRS basis is the spread between the cross-currency swap (CCS) rate and interest rate swap (IRS) rate and is often used as an indicator for foreign exchange soundness along with the CDS premium in Korea. Based on the phenomenon in which the USDKRW exchange rate rises (falls) when the CCS-IRS basis widens (tightens), this study proposes to use the CCS-IRS basis to build a foreign exchange trading model of the USDKRW exchange rate. The experimental results show that our NNs trading model of the USDKRW exchange rate outperforms XGBoost and LR model, generating an annual return of 4.888%∼8.464%, whereas the buy-and-hold and sell-and-hold strategies in the USDKRW exchange rate market produces an annual return of 2.2% and −1.15%, respectively. The foreign exchange trading model established in this paper not only enables foreign exchange traders to obtain sustainable economic profits, but also contributes to the efficiency of the financial market, thereby contributing to sustainable economic growth. [ABSTRACT FROM AUTHOR]

• We suggest a method for automatically finding potential business partners. • Deep learning is applied to capture transaction patterns between companies. • The proposed model is effective to identify potential business partners. • Recommendation performances of the proposed model superior to conventional ones. Potential business partner (BP) recommendation is one of the most important issues for companies to increase sales opportunities by discovering new candidate buyers. Recommendation at a low cost and automation is especially essential for small and medium businesses. However, identifying potential BPs has been regarded as a challenging task since not only an analysis of business characteristics for each candidate company by human experts is required but also an investigation of all the possible combinations of their matchings is necessary. Therefore, in this paper, we propose novel BP recommendation models, called deep business partner recommendation (DBR) models, that aim to automatically suggest potential BPs. Specifically, deep learning technique is applied to understand hidden transaction patterns between companies with various industrial sectors and product properties through the two-phases involving i) BP relation representation phase and ii) training and testing phase. In the former, for each company, its features including the industrial sector, product property, relative transaction volume, and geographical distance are embedded into a vector for utilizing as the input of the proposed models. In the latter, the proposed DBR models repeatedly use the input values to capture the hidden transaction patterns between companies in the training. In the testing phase, the extensive experiments conducted to evaluate the BP recommendation performances of the proposed DBR models using a real-world dataset consisting of transaction records among companies in South Korea. The experiment results show that the suggested DBR models significantly outperform the conventional models, in terms of the accuracy for the BP recommendation tasks. [ABSTRACT FROM AUTHOR]

Abstract: The financial distress forecasting has long been of great interest both to scholars and practitioners. The financial distress forecasting is basically a dichotomous decision, either being financial distress or not. Most statistical and artificial intelligence methods estimate the probability of financial distress, and if this probability is greater than the cutoff value, then the prediction is to be financial distress. To improve the accuracy of the financial distress prediction, this paper first analyzed the yearly financial data of 1888 manufacturing corporations collected by the Korea Credit Guarantee Fund (KODIT). Then we developed a financial distress prediction model based on radial basis function support vector machines (RSVM). We compare the classification accuracy performance between our RSVM and artificial intelligence techniques, and suggest a better financial distress predicting model to help a chief finance officer or a board of directors make better decision in a corporate financial distress. The experiments demonstrate that RSVM always outperforms other models in the performance of corporate financial distress predicting, and hence we can predict future financial distress more correctly than any other models. This enhancement in predictability of future financial distress can significantly contribute to the correct valuation of a company, and hence those people from investors to financial managers to any decision makers of a company can make use of RSVM for the better financing and investing decision making which can lead to higher profits and firm values eventually. [Copyright &y& Elsevier]

Abstract: The aim of this paper is to present and test a model of innovativeness improvement based on the impact of organizational learning culture. The concept of organizational learning culture (OLC) is presented and defined as a set of norms and values about the functioning of an organization. They should support systematic, in-depth approaches aimed at achieving higher-level organizational learning. The elements of an organizational learning process that we use are information acquisition, information interpretation, and behavioral and cognitive changes. Within the competing values framework OLC covers some aspects of all four different types of cultures: group, developmental, hierarchical, and rational. Constructs comprising innovativeness are innovative culture and innovations, which are made of technical (product and service) and administrative (process) innovations. We use data from 201 Korean companies employing more than 50 people. The impact of OLC on innovations empirically tested via structural equation modeling (SEM). The results show that OLC has a very strong positive direct effect on innovations as well as moderate positive indirect impact via innovative culture. [Copyright &y& Elsevier]

Abstract: In a bankruptcy prediction model, the accuracy is one of crucial performance measures due to its significant economic impact. Ensemble is one of widely used methods for improving the performance of classification and prediction models. Two popular ensemble methods, Bagging and Boosting, have been applied with great success to various machine learning problems using mostly decision trees as base classifiers. In this paper, we propose an ensemble with neural network for improving the performance of traditional neural networks on bankruptcy prediction tasks. Experimental results on Korean firms indicated that the bagged and the boosted neural networks showed the improved performance over traditional neural networks. [Copyright &y& Elsevier]

Abstract: The telecommunications industry is undergoing dramatic changes fueled by rapid technical development and regulatory changes. Especially, in Korea, the penetration rate of mobile service is almost 90%; therefore, there is not enough room for a new customer. Furthermore, today’s new services have been influenced by more factors in decision making before launching. It has been the main issue for decision makers whether a new service will be successful or not. However, a few methods have been introduced to evaluate the feasibility of a new service. And very few papers have introduced to the evaluatation of comparative advantages among competitive services, but such methods mainly depend on subjective opinion of experts. The proposed model considers the multidimensional factors influencing telecom service evaluation based on the ANP’s advantages, which helps service providers in a telecom field to select a new service with the view of their strategies. Also, the BSC perspectives are adapted in the model to generate evaluation measures. Additionally, a new video phone service in Korea is evaluated based on a proposed model. [Copyright &y& Elsevier]

Abstract: In Korea, information infrastructure industry has been expanding its business. However, service level of information infrastructure has not reached the expected level of customers in the context of CRM (Customer Relationship Management) yet. In this paper, we develop a structural equation model for Customer Satisfaction Index (CSI) to measure the quality level of information infrastructure reflecting the aspects of information infrastructure industry. Then, we compare the KII (Korea Information Infrastructure) satisfaction Indices from various customer groups using the structural equation model (SEM). It is expected that the results of our study suggest helpful strategies to increase the satisfaction level of KII customers. [Copyright &y& Elsevier]

Abstract: For the effective operation of the air power in the modern war, aircraft should be well-maintained by preventing a series of failures. However, the current maintenance system employed by ROKAF (Republic of Korea Air Force) does not fully utilize cumulative sequential failure data. In this paper, we apply sequential association rules to extract the failure patterns and forecast failure sequences of ROKAF aircrafts according to various combinations of aircraft types, location, mission and season. It is expected that our analysis can add value to the existing maintenance database. Also, our approach can improve the utilization of aircrafts by properly forecasting the future demand of aircraft spare parts. [Copyright &y& Elsevier]

The experience of Singapore and South Korea makes it clear that under certain circumstances massive testing is an effective way for containing the advance of the COVID-19. In this paper, we propose a modified SEIR model which takes into account tracing and massive testing, proving theoretically that more tracing and testing implies a reduction of the total number of infected people in the long run. We apply this model to the spreading of the first wave of the disease in Spain, obtaining numerical results. After that, we introduce a heuristic approach in order to minimize the COVID-19 spreading by planning effective test distributions among the populations of a region over a period of time. As an application, the impact of distributing tests among the counties of New York according to this method is computed in terms of the number of saved infected individuals. • A SEIR model is introduced to analyse the efficiency of test distributions. • It is proved that massive testing helps reducing the number of infected people. • A heuristic method is developed to increase the number of saved infections. • Extensive computational experience quantifies the number of saved infections. [ABSTRACT FROM AUTHOR]

• We identify historical factions using multi-layered network of historical figures. • Multi-layered network of historical figures is constructed using family relations. • Semi-supervised learning for multi-layered networks is used to identify the factions. • The identification benefits from matrix sparseness incurred from historical data. • Experiments show that the proposed method can provide meaningful insights on history. With immense influx of historical data, quantitative inferences on history based on machine learning is becoming more prevalent, attracting many researchers. In particular, understanding the dynamics of historical factions is important as they shared academic beliefs, political views and interests, in which the interactions between the factions portray general political, social, and economic structure of a certain era. In recent years, studying such dynamics through network-based methods on human networks, constructed from genealogy data, have shown promising results. In this paper, we enhance the identification of historical factions by exploiting multi-layered network of historical figures. To understand the mechanisms of historical factions, it is pivotal to comprehend the change in relation between important historical events. The proposed method consists of constructing a multi-layered network of historical figures and applying semi-supervised learning framework to identify historical factions. The proposed method was applied to the classification of factions in the political turmoil occurred during the 15th to 16th century Korea. [ABSTRACT FROM AUTHOR]