• Journal of Intelligence and Information Systems
• /
• v.29 no.2
• /
• pp.241-265
• /
• 2023

A corporate insolvency prediction model serves as a vital tool for objectively monitoring the financial condition of companies. It enables timely warnings, facilitates responsive actions, and supports the formulation of effective management strategies to mitigate bankruptcy risks and enhance performance. Investors and financial institutions utilize default prediction models to minimize financial losses. As the interest in utilizing artificial intelligence (AI) technology for corporate insolvency prediction grows, extensive research has been conducted in this domain. However, there is an increasing demand for explainable AI models in corporate insolvency prediction, emphasizing interpretability and reliability. The SHAP (SHapley Additive exPlanations) technique has gained significant popularity and has demonstrated strong performance in various applications. Nonetheless, it has limitations such as computational cost, processing time, and scalability concerns based on the number of variables. This study introduces a novel approach to variable selection that reduces the number of variables by averaging SHAP values from bootstrapped data subsets instead of using the entire dataset. This technique aims to improve computational efficiency while maintaining excellent predictive performance. To obtain classification results, we aim to train random forest, XGBoost, and C5.0 models using carefully selected variables with high interpretability. The classification accuracy of the ensemble model, generated through soft voting as the goal of high-performance model design, is compared with the individual models. The study leverages data from 1,698 Korean light industrial companies and employs bootstrapping to create distinct data groups. Logistic Regression is employed to calculate SHAP values for each data group, and their averages are computed to derive the final SHAP values. The proposed model enhances interpretability and aims to achieve superior predictive performance.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2020.06a
• /
• pp.21-22
• /
• 2020

낙동강 하굿둑(이하 하굿둑)은 1987년 부산 사하구와 강서구 사이에 건설되어 하류 지역의 바닷물 유입을 막아 부산, 울산, 경남 등에 안정적으로 생활·농업·공업 등의 분야에 용수를 공급하는 역할을 해왔다. 현재, 하굿둑의 수문은 낙동강 상류로부터 하류로 흘러내려오는 민물(담수)을 방류하기 위해서만 하굿둑 수문을 개방하고 있다. 하구는 하천의 담수와 바다의 염수가 서로 만나는 구역으로 바닷물과 염수의 밀도차에 의한 혼합으로 자연상태의 하구에서는 담수와 염수가 섞이는 기수역이 형성되며, 이러한 특성으로 하구 인근의 지역에서는 일반적인 하천 및 해양, 연안과는 분명히 구별되는 생태계가 조성된다. 하굿둑 건설이후 바닷물(해수)과 민물(담수)이 만나는 낙동강 어귀에 기수생태계가 사라지면서 바닷물이 유입될 수 있도록 하여 생태계를 복원해야 한다는 필요성이 제기되어 왔으며, 하굿둑이 지역에 기여해온 사실은 분명하나 하굿둑으로 인해 생태계 단절이 발생하고 기수생태계가 파괴되었기 때문에 이를 해결하기 위해서는 하굿둑을 개방하여 과거 기수생태계를 복원해야 한다는 목소리가 높아지고 있다. 이에 따라 정부에서는 하굿둑의 기수생태계 복원을 위해서 관계기관 합동으로 의사결정을 하고 효율적인 개방 방안을 모색하는 실무협의회를 구성하여 운영 중이고, 실무협의회 논의를 통해 5개 주요 관계기관(환경부, 국토부, 해양수산부, 부산광역시, K-water) 공동으로 "낙동강하굿둑 운영개선 및 생태복원 방안 연구용역"을 추진 중이다. 2018년 1단계 용역이 완료되었으며, 2019년부터 2단계 연구용역을 추진 중이고 하굿둑 개방의 수준별로 각종 영향을 검토한 후 대책을 마련하여 기수생태계 복원 방안을 수립하는데 그 목적이 있다. 2단계 연구용역에서는 과학적이고 합리적인 기수생태계 복원방안 마련을 위해서 실제로 해수를 유입시키는 3차례의 실증실험 및 수리모형실험 등을 추진한다. 기존 연구들에서도 수문개방에 따른 해수유입 영향에 대해 모델링을 통해서 분석했지만 이는 검증이 이루어지지 않은 결과로 이번 용역에서는 실제 해수를 유입시키고 염분의 침투 및 각종 수생태 영향을 모니터링 한 후 그 결과를 반영하여 모델링을 고도화하고 있다. 최종적으로 고도화된 모델링 결과를 기반으로 기수생태계 조성 방안별로 염분, 수질, 수생태, 침퇴적 등 각종 분야에 대한 정확한 영향을 분석하고 이에 대한 대책을 포함하여 최종적으로 바람직한 기수생태계 복원 방안을 제시할 계획이다. 기수생태계 복원 방안이 계획에만 그치지 않고 실행으로 연결시키기 위해서 필요성에 대한 이해를 바탕으로 공감대를 형성해 나아가고 있으며 지역주민, 전문가, 관계기관 등 민(民)·관(官)·학(學) 다양한 의견을 수렴하여 하구지역내 수량-수질-수생태를 종합적으로 고려하여 복원 방안을 마련 후 사회적인 합의를 추진하여 확정할 예정이며, 하구의 안정적인 관리를 위해 AI 등 4차 산업혁명기술을 적극 적용하는 스마트한 하구물관리(Smart Estuary Watershed Management)"를 활용한 "하구통합물관리" (Estuary Integrated Watershed Management) 등 과학적인 관리를 추진할 계획이다.

• Smart Media Journal
• /
• v.12 no.10
• /
• pp.55-62
• /
• 2023

Innovation and change are occurring rapidly in the agriculture and livestock industry, and new technologies such as smart bams are being introduced, and data that can be used to control equipment is being collected by utilizing various sensors. However, there are various challenges in the operation of bams, and virtual sensor technology is needed to solve these challenges. In this paper, we define various data items and sensor data types used in livestock farms, study cases that utilize virtual sensors in other fields, and implement and design a virtual sensor system for the final smart livestock farm. MBE and EVRMSE were used to evaluate the finalized system and analyze performance indicators. As a result of collecting and managing data using virtual sensors, there was no obvious difference in data values from physical sensors, showing satisfactory results. By utilizing the virtual sensor system in smart livestock farms, innovation and efficiency improvement can be expected in various areas such as livestock operation and livestock health status monitoring. This paper proposes an innovative method of data collection and management by utilizing virtual sensor technology in the field of smart livestock, and has obtained important results in verifying its performance. As a future research task, we would like to explore the connection of digital livestock using virtual sensors.

• Journal of Venture Innovation
• /
• v.6 no.1
• /
• pp.59-73
• /
• 2023

In the era of the 4th Industrial Revolution, manufacturing is the implementation of smart factories through digital transformation, and refers to consumer-centered intelligent factories that combine next-generation digital new technologies and manufacturing technologies beyond the existing factory automation level. In order to successfully settle such a smart factory, it is necessary to train professionals. However, education for smart factories is difficult to have actual field mechanical facilities or overall production processes. Therefore, there is a need for a system that can visualize and control the flow and process of logistics at the actual production site. In this paper, the logistics flow of the actual site was implemented as a small FMS, a physical system, and the production process was implemented as a digital system. In real-time synchronization of the physical system and the digital system, the location of AGV and materials, and the process state can be monitored to see the flow of logistics and process processes at the actual manufacturing site. The developed digital twin system can be used as an effective educational system for training manpower in smart factories.

• Korean Journal of Agricultural and Forest Meteorology
• /
• v.25 no.3
• /
• pp.182-196
• /
• 2023

Accurate and timely estimation of crop yields is crucial for various purposes, including global food security planning and agricultural policy development. Remote sensing techniques, particularly using vegetation indices (VIs), have show n promise in monitoring and predicting crop conditions. However, traditional VIs such as the normalized difference vegetation index (NDVI) and enhanced vegetation index (EVI) have limitations in capturing rapid changes in vegetation photosynthesis and may not accurately represent crop productivity. An alternative vegetation index, the near-infrared reflectance of vegetation (NIRv), has been proposed as a better predictor of crop yield due to its strong correlation with gross primary productivity (GPP) and its ability to untangle confounding effects in canopies. In this study, we investigated the potential of NIRv in estimating crop yield, specifically for corn and soybean crops in major crop-producing regions in 14 states of the United States. Our results demonstrated a significant correlation between the peak value of NIRv and crop yield/area for both corn and soybean. The correlation w as slightly stronger for soybean than for corn. Moreover, most of the target states exhibited a notable relationship between NIRv peak and yield, with consistent slopes across different states. Furthermore, we observed a distinct pattern in the yearly data, where most values were closely clustered together. However, the year 2012 stood out as an outlier in several states, suggesting unique crop conditions during that period. Based on the established relationships between NIRv peak and yield, we predicted crop yield data for 2022 and evaluated the accuracy of the predictions using the Root Mean Square Percentage Error (RMSPE). Our findings indicate the potential of NIRv peak in estimating crop yield at the county level, with varying accuracy across different counties.

• The Journal of the Convergence on Culture Technology
• /
• v.9 no.5
• /
• pp.867-874
• /
• 2023

Recently, many problems are occurring in the bee ecosystem due to rapid climate change. The decline in the bee population and changes in the flowering period are having a huge impact on the harvest of bee-keepers. Since it is impossible to continuously observe the beehives in the hive with the naked eye, most people rely on knowledge based on experience about the state of the hive.Therefore, interest is focused on smart beekeeping incorporating IoT technology. In particular, with regard to swarming, which is one of the most important parts of beekeeping, we know empirically that the swarming time can be determined by the sound of the queen bee, but there is no way to systematically analyze this with data.You may think that it can be done by simply recording the sound of the queen bee and analyzing it, but it does not solve various problems such as various noise issues around the hive and the inability to continuously record.In this study, we developed a system that records queen bee sounds in a real-time cloud system and analyzes sound patterns.After receiving real-time analog sound from the hive through multiple channels and converting it to digital, a sound pattern that was continuously output in the queen bee sound frequency band was discovered. By accessing the cloud system, you can monitor sounds around the hive, temperature/humidity inside the hive, weight, and internal movement data.The system developed in this study made it possible to analyze the sound patterns of the queen bee and learn about the situation inside the hive. Through this, it will be possible to predict the swarming period of bees or provide information to control the swarming period.

• Journal of the Korean Geotechnical Society
• /
• v.40 no.3
• /
• pp.85-91
• /
• 2024

To realize stable use of ground treated with cementitious materials, the curing process must be evaluated. In this study, a time domain reflectometry (TDR) measurement system was employed to evaluate the curing process of cementitious grout based on the electromagnetic property. A coated probe was manufactured to prevent electrical connection between the electrodes by the electrically conductive cementitious grout, and a calibration process was performed to estimate the actual relative permittivity using the coated probe. To assess the curing process of cementitious grout using the TDR measurement system, cementitious grout with added retarder was prepared with a water-to-cement ratio of 45%. A preliminary measurement was conducted immediately after pouring the cementitious grout into the mold to test the applicability of the coated probe, and TDR signals and relative permittivity were measured at 3~288 hours of curing time. The experimental results demonstrate that the relative permittivity of the cementitious grout immediately after pouring was greater than 100, decreased rapidly over time, and converged to approximately 13.8 at 144 hours, which is considered the fully cured time. This findings of this study demonstrate that the TDR measurement system with a coated probe is applicable to electrically conductive materials. In addition, the TDR measurement system can be used effectively to monitor the curing process of cementitious grout based on electromagnetic properties.

• Economic and Environmental Geology
• /
• v.57 no.2
• /
• pp.271-280
• /
• 2024

This study examines the European Union (EU)'s policies on managing nitrate contamination in groundwater and provides implications for the future groundwater management in South Korea. Initiated by the 1991 Nitrate Directive, the EU has pursued a multifaceted approach to reduce agricultural nitrate pollution through sustainable ('good') farming practices, regular nitrate level monitoring, and designating Nitrate Vulnerable Zones. Further policy integrations, like the Water Framework Directive and Groundwater Directive, have established comprehensive protection strategies, including the use of pollutant threshold values. Recently, the 2019 Green Deal escalated efforts against nitrates, aligning with broader environmental and climate objectives. This review aims to explore these developments, highlighting key mitigation strategies against nitrate pollution, and providing valuable insights for the future sustainable groundwater nitrate management in South Korea, emphasizing the importance of preventive measures and collaborative efforts to restore and improve groundwater quality.

• Korean Journal of Environment and Ecology
• /
• v.38 no.3
• /
• pp.310-323
• /
• 2024

The purpose of this study was to understand the vegetation structure and ecological characteristic of Bulgapsan(Mt.) Provincial Park by setting up and surveying 64 plots(100m²). The analysis using the TWINSPAN and DCA techniques found seven community groups: Pinus densiflora-Quercus variabilis community, P. densiflora-P. rigida-Q. serrata community, Q. variabilis-Carpinus tschonoskii community, Q. aliena-Q. variabilis-Cornus controversa community, Q. aliena-Platycarya strobilacea community, Broad-leaved miced community and Q. variabilis community. The result of vegetation community structure analysis showed that P. densiflora community and deciduous Quercus spp. community were in competition, and succession to Quercus spp. community was expected. In the case of other broad-leaved forests, the current status is expected to be maintained. But continuous monitoring is required in areas where Neolitsea sericea and Cephalotaxus appear, which grow naturally in warm temperate forest and southern temperate vegetation zone. Species diversity by communities are confirmed to be highest at 2.6654 in the actively competitive P. densiflora-P. rigida-Q. serrata community, and the lowest in the Deciduous broad-leaved forests community at 1.2548. The results of the tree rings and annual growth analysis showed that dominant trees had an average age of more than 37~87 years. Among them, N. sericea designated as a natural monument was 48~56 years old.

• Korean Chemical Engineering Research
• /
• v.62 no.3
• /
• pp.238-245
• /
• 2024

The lactate electrode can be utilized either as an electrode for lactate sensor to monitor the patient's health status, stress level, and athlete's fatigue in real time or lactate fuel cell. In this study, we fabricated a high-performance electrode composed of lactate oxidase, catalase, and mitochondria, and investigated the surface analysis and electrochemical properties of this electrode. Carbon paper modified with single-walled carbon nanotubes (CP-SWCNT) had significantly improved electrical conductivity compared to before modification. The electrode to which lactate oxidase, catalase, and mitochondria were attached (CP-SWCNT-LOx-Cat-Mito) produced a higher current than the electrode to which lactate oxidase and catalase were attached. The amount of reduction current produced by the bilirubin oxidase (BOD)-attached electrode (CP-SWCNT-BOD) was greatly affected by the presence or absence of oxygen in the electrolyte. The fuel cell composed of CP-SWCNT-LOx-Cat-Mito (anode) and CP-SWCNT-BOD (cathode) produced maximum power (29 ㎼/cm²) at a discharge current density of 133 ㎂/cm². From this study, we had proved that mitochondria is essential for improving lactate sensor and fuel cell performance.