An Ensemble Based Default Forecasting Model for Economic Payoff Maximization

Seungyoo Jeon,Chan Park,Kisung Yang 한국재무학회 2023 한국재무학회 학술대회 Vol.2023 No.11

This study addresses default prediction through ensemble techniques and cost-sensitive learning. It introduces a novel stacking method, focusing on instances with varying misclassification costs. Previous research lacks this comprehensive approach, highlighting a research gap. The proposed technique proves advantageous in terms of economic payoff and performance, offering practical utility. It complies with regulatory monitoring standards without incurring statistical cost penalties. This study demonstrates potential revenue gains even without precise cost ratios. Empirical results using Taiwanese company bankruptcy data (1999-2009, 95 financial ratios) reveal significant outcomes. Firstly, the proposed algorithm markedly improves economic payoff. Secondly, its statistical performance remains unaffected, even considering the dependency on overall prediction errors related to misclassifying defaults. Lastly, the method is computationally efficient, robust across cost scenarios.

Strategies to improve the photocatalytic activity of TiO₂: 3D nanostructuring and heterostructuring with graphitic carbon nanomaterials

Lee, Kisung,Yoon, Hyewon,Ahn, Changui,Park, Junyong,Jeon, Seokwoo The Royal Society of Chemistry 2019 Nanoscale Vol.11 No.15

<P>TiO2-based photocatalysis has been considered to be one of the most promising avenues for environmental remediation including water purification. However, several technical issues such as the limited surface area of bulk TiO2, the large band gap of TiO2, and rapid charge recombination still limit the practical application of TiO2 photocatalysts. Therefore, here we focus on two structural design strategies: (i) monolithic three-dimensional (3D) nanostructuring, and (ii) heterostructuring with graphitic carbon nanomaterials. A monolithic 3D nanostructure enables maximal surface area in a given volume and efficient reuse of the photocatalyst without recollection. Heterostructuring with carbon nanomaterials helps achieve maximal utilization of the solar spectrum and charge separation and provides efficient TiO2 photocatalysts. In this review, recent progress on TiO2 photocatalysts toward the abovementioned strategies will be summarized. Further discussion and direction will provide insights into the rational design of highly efficient TiO2 photocatalysts, and help develop advanced photocatalyst models.</P>

Default Prediction Modeling based on economic costs Minimization

Chan Park,Seungyoo Jeon,Kisung Yang 한국재무학회 2023 한국재무학회 학술대회 Vol.2023 No.11

In the default prediction problem, thecost from thefailure of forecasting defaults is much biggerthan that of forecastingnon-defaults. Thecost asymmetryis deeperin the corporatedefaultpredictionthan theretailas corporateloan portfolios arenot granular. However, thetwo typesof costs aretreatedequallyin generalas defaultpredictionmodels areusually estimatedto minimizepredictionerrorsor maximizestatistical performance. This practicemight not fulfill thegoal of risk managementto minimizeeconomiclosses. To mitigatethis issue, this study apply cost-sensitivelearningapproach to defaultprediction, which minimizeseconomiccosts insteadof statistical errors. Wedefineeconomic costs and testthemfor various levelsof thecost asymmetryby employingLogistic regression, XGBoost, and LightGBM. As a resultof empiricalexperimentswith Taiwaneseand Polish corporate default data, we first find that the proposed cost-sensitivemodels are superiorto thecost-insensitivecounterpartsin termsof economiccost, mostly regardless of thecost asymmetryscenarios. Secondly, neverthele,ssthedecreasesin thestatistical performanceare relativelysmall – economic costs decrease24.6% at theexpenseof the decrease in AUC of 4.6% on average. This suggests that financial firms can adopt the proposed default prediction models without violating the regulatory requirement on model quality. Lastly, we find that the features of high prediction power in the cost-sensitive and insensitive models are different, which has an important implication for credit monitoring.

Comparison of the Microstructural Characterizations of GaN Layers Grown on Si (111) and on Sapphire

Hui-Youn Shin,Kisung Jeon,Youngil Jang,Mingu Gang,Myungshin Choi,박원화,박규호 한국물리학회 2013 THE JOURNAL OF THE KOREAN PHYSICAL SOCIETY Vol.63 No.8

Due to the large differences in the lattice constants and the thermal expansion coefficients betweenGaN and Si, GaN growth on a Si substrate usually leads initially to high defect densities andcracks. If high-quality GaN films on Si substrate are to be obtained, it is essential to understandthe different growth characteristics of GaN layers grown on Si and on sapphire. In this study, theGaN specimens were grown on sapphire and Si (111) substrates with AlGaN and AlN buffer layers,respectively, by metalorganic chemical vapor deposition (MOCVD). Using transmission electronmicroscopy (TEM) and micro-Raman spectroscope, we carried out a comparative investigation ofGaN growth by characterizing lattice coherency, defect density, and residual strain. These analysesrevealed that the GaN layers grown on Si have much residual tensile strain and that strain has aneffect on the formation of InGaN/GaN multiple quantum wells (MQWs) above the GaN layers.

Default Prediction Modeling based on economic costs Minimization

Chan Park,Seungyoo Jeon,Kisung Yang 한국재무학회 2023 한국재무학회 학술대회 Vol.2023 No.11

In the default prediction problem, the cost from the failure of forecasting defaults is much bigger than that of forecasting non-defaults. The cost asymmetry is deeper in the corporate default prediction than the retail as corporate loan portfolios are not granular. However, the two types of costs are treated equally in general as default prediction models are usually estimated to minimize prediction errors or maximize statistical performance. This practice might not fulfill the goal of risk management to minimize economic losses. To mitigate this issue, this study apply cost-sensitive learning approach to default prediction, which minimizes economic costs instead of statistical errors. We define economic costs and test them for various levels of the cost asymmetry by employing Logistic regression, XGBoost, and LightGBM. As a result of empirical experiments with Taiwanese and Polish corporate default data, we first find that the proposed cost-sensitive models are superior to the cost-insensitive counterparts in terms of economic cost, mostly regardless of the cost asymmetry scenarios. Secondly, nevertheless, the decreases in the statistical performance are relatively small – economic costs decrease 24.6% at the expense of the decrease in AUC of 4.6% on average. This suggests that financial firms can adopt the proposed default prediction models without violating the regulatory requirement on model quality. Lastly, we find that the features of high prediction power in the cost-sensitive and insensitive models are different, which has an important implication for credit monitoring.

<i>N</i> -Glycosylation Modification of Plant-Derived Virus-Like Particles: An Application in Vaccines

Kim, Hyun-Soon,Jeon, Jae-Heung,Lee, Kyung Jin,Ko, Kisung Hindawi Publishing Corporation 2014 BioMed research international Vol.2014 No.-

<P>Plants have been developed as an alternative system to mammalian cells for production of recombinant prophylactic or therapeutic proteins for human and animal use. Effective plant expression systems for recombinant proteins have been established with the optimal combination of gene expression regulatory elements and control of posttranslational processing of recombinant glycoproteins. In plant, virus-like particles (VLPs), viral “empty shells” which maintain the same structural characteristics of virions but are genome-free, are considered extremely promising as vaccine platforms and therapeutic delivery systems. Unlike microbial fermentation, plants are capable of carrying out <I>N</I>-glycosylation as a posttranslational modification of glycoproteins. Recent advances in the glycoengineering in plant allow human-like glycomodification and optimization of desired glycan structures for enhancing safety and functionality of recombinant pharmaceutical glycoproteins. In this review, the current plant-derived VLP approaches are focused, and <I>N</I>-glycosylation and its in planta modifications are discussed.</P>

금속제 기구 및 용기포장의 유해물질에 관한 조사 연구 : 금속제 주방 기구에서 유해중금속 용출에 관한 연구 Study on Migration of Trace Elements from Metallic Kitchenwares

이광호,권기성,곽인신,전대훈,최병희,유승석,김성욱,이선희 식품의약품안전청 1998 식품의약품안전청 연보 Vol.2 No.-

구이웅즉불판, 스데"1크용 불판, 전골그릇. 솥, 프라이괜, 갬비, 주전자, 밥그릇 및 석쇠 둥 주방용 기구에 사웅되는 황동(3'3총)· 묵쇠 (17종), 알루미늄(10종), 청동(2종), 구리 (2종), 스테인레스(2종) 둥의 금속제에서 Pb, Cd, Ct Zn, Sn등 유괘 중금속의 용출과 기작에 대하여 연구하였다. 촹동을 대표금속으로 선정하여 용출 온도(30, 60, 80,95'c), PH(2.5, 4.3, 6.0, 7.0) 및 시간(30, 90, 180, 3:60븐)에 딱른 금속 용출량 조사에서 온도 및 P 변화체 딱른 시험결과, 95'C, pH 2.5인 조건에서 용출이 가장 많았던 반면, 용출 시간별 시험에서는 180분 이후까지 Pb가 용출된 후 편형 상태를 유지하였다. 위의 실험 결과로 미루어 현행 식품공전 시험법인 30분 조건에 대한 추후 재검토 연구가 필요하다고 잔단되며, 납이 용출되는 기작은 SEM을 통하띨 확인항 수 있었다. 황동의 경우, 재질줄의 Pb 항량곽용 출량과의 관계를 규명하기 위해 Pb 함략별 10종(0, 0.02, 0.09, 0.1, 0.2, 0.5, 0-9, 1.7, 5.4, 9.2%)의 시껼를 제작하여 4% 초산, 95'C, 30분에서 시험한 결과, 용출규격인 중금속이 1.Oppm(납으로서) 이하가 되기 위해서는 재질중 허용 Pb 함량은 0.2% 이하가 되어야 했다. 시중에 운통중인 각 재질별 금속제 기구에서 용출되는 유해중금속의 실태를 파악한결과, 황동제의 경우 주로 주물제품의 구이용 불판(양식 스테이크용 포함)으로 20종중 17종인 식풍 공전의 기준 ·규격인 1.OPPm이상의 Pb가 검출되어 85%의 련적합율을 뽀였다. Pb 용출 함량별 분포를 보면 1~10ppin이 15%, 20~100ppmo1 30%, 100~200ppmo1 20%, foo~300ppmol 10%, 300~400ppmo1 5%, 400~500ppr1 이 5%로 10~100ppm 사이의 용출량을 나타내는 제품이 가장 많이 유통되는 것으로 확인 되었다. 그외 Cu7l 0.37~77.85ppm, Zinc가 2.88~221.Oeppm, Fe이 미량 용출되었으며 Cd는 검출되지 않았다. 무꼭는 요즘 구이용 불판으로 사용되고 있는 솥뚜껑r 스테이크용 불판, 솥, 전골냄티, 프라이팬 등을 대상 겋체로 하였다. Fe이 0.S2~5,058 ppin, Cu, Zn 및 Pb등이 미량 검출 되었으며 Sn, Cd은 용출되지 않아 기준 ·규격에 적합한 것으로 판단.되었다. 알루디늄제는 램비 및 주전짜를 대상 검체로 하였는데 Cu, Zn, Sn, 및 Pb 등이 식품공전의 기준 · 규격의 IPPU내로 미량 검출되엇으며 Cd은 검출되지 않아 역시 기준·규격에 적합하다는 걸론을 얻었다. The migration of trace elements, Pb, Cd, Cu, Zn and Sn from metallie food-contactutensiles was investigated. All samples used in this work were grills, cooking pots, kettles, frying pansand bowls made of brass(20 samples), iron(17 samples) , aluminium(10 samples), copper(2 samples),bronze(2 samples) or stainless(2 samples). The migration of metats from brasses was affected by the iai-gration temperature(30, 60, 80 and 95'c ), the simulant pH (2.5, 4.3, 6.0 and 7.0) and the migration tirae(30,90, 180 and 360mia.). T,he amount of Pb migrati,oB was maximum at 95'c with pH 2.5. This study in-dicated that it could be required to reexamine the m igration time as 30 minutes suggested by the KoreaFood Code because the equilibrium of Pb migration was reached at 180 minutes. The mechanism of Pbmigration was investigated with Scanning Electron :Microscope(SEM) and Electron Probe Micro Analya-er(EPMA ). To correlate the relations between the m igration and the content of Pb in brass samples, tinbrass samples having known eontent of Pb(the quaritities of Fb in brass ; 0, 0.02, 0.09, 0,3,0.2, 0.5,0.9,1.4, 5.4 and 9.2% (w/w) ) were tested in the following conditions,4% acetic acid,95·C and 30 minutes. Theresult re?i?seBts that content of Pb in brass samples must be below 0.2% (w/w) to satis(y the 1.0 ppm(as Pb), of forea Food Code. The study was also coBlducted to estimate the migration of harmful metalsin metallic kitchenware made of brass, iron, alumiBium, copper, bronze and stainless. In the case ofbrass samples, the migration of Pb exceeded the 1.0 ppm in 17 among 20 samples with 85% rejectionrate. The pattern of the distribution for Pb migratioB was as followed; 15% : 1~10ppn1; 30% : fO~100ppm; 20% : 100~200ppm; 10% : 200~300ppm, 5% : 300~400ppm; 5% : 400~500ppm, respectivelr,while the content of otller metals detected as Cu 0.3,7~77.85ppm, Zn 2.88~221.Ooppm, Fe trace. Variousiron kitchenwares including sTilts, cooking pots ancl frying pans were also tested. The migration of themetals showed the following result, Fe . 0.52~s.07 ppm; Cu, Zn, Pb : trace; Sn. Cd : not detected.respectively. Aluminium kitchenwares like kettles were tested. There was no significant migration of trace metals, Cu, Zn, Sn, Fe and Pb, meanwhile the migration of Cd in samples made of the aluminium products could not be detected. The results demonstrated that iron and aluminium kitchenwares were suitable for Korea Food Code, however, brass products could provide harmful effect on human health.

3D CT-Fluoroscopy 시스템을 위한 고성능 디지털 X-선 센싱 기술개발

허영(Young Huh),전성채(Sungchae Jeon),이동훈(Dong Hoon Lee),김봉희(Bonghoe Kim),황윤금(Yoon Geum Hwang),권오경(Oh-Kyong Kwon),이기성(Kisung Lee),박문석(Moon-Seok Park),정진엽(Chin Youb Chung) 대한전자공학회 2010 대한전자공학회 학술대회 Vol.2010 No.6

Expression of Aβ-Fc Fusion Protein in Transgenic Potato

Kim, Hyun-Soon,Youm, Jung Won,Lee, Jeong-Hwan,Jeon, Jae-Heung,Ko, Kisung Korean Society of Horticultural Science 2014 원예과학기술지 Vol.32 No.3

Transgenic potato was generated to express recombinant 5 repeated ${\beta}$-amyloid ($A{\beta}$) peptides, potential antigens to be applied as a preventive accine for Alzheimer's disease using Agrobacterium mediated transformation. The $A{\beta}$ peptides were fused to the human IgG Fc fragment enhancing protein and KDEL, which is the endoplasmic reticulum (ER) retention signal ($5A{\beta}$-FcK). The $5A{\beta}$-FcK, was expressed under the control of the duplicated 35S promoter. PCR analysis confirmed the presence of the transgene in several transgenic potato lines. Southern blot analysis showed only a single gene copy number in transgenic line 22, whereas multiple gene copy numbers were shown for transgenic lines 31 and 44. Northern blot analysis showed that line 22 had stronger mRNA levels when compared to lines 31 and 44. Immunoblot analysis confirmed that the $5A{\beta}$-FcK protein was expressed in the transgenic potato plant. These results indicate that $5A{\beta}$ fused to Fc can be expressed in potato plants.

Symmetry dictated grain boundary state in a two-dimensional topological insulator

Seoung-Hun Kang,Hyo Won Kim,Hyun-Jung Kim,Kisung Chae,Suyeon Cho,Wonhee Ko,Sangjun Jeon,Se Hwang Kang,HeejunYang,Sung Wng Kim,Young-Kyun Kwon,Young-Woo Son 한국진공학회 2020 한국진공학회 학술발표회초록집 Vol.2020 No.8