GA-based Normalization Approach in Back-propagation Neural Network for Bankruptcy Prediction Modeling

Qiu-yue Tai(태추월),Kyung-shik Shin(신경식) 한국지능정보시스템학회 2010 지능정보연구 Vol.16 No.3

역전파 알고리즘은 오랫동안 부도예측모형 관련한 연구에 많이 적용되어왔다. 역전파 알고리즘을 사용하기전에 필히 고려해야 할 중요한 요소들로는 네트워크 구조, 학습요소, 정규화 방법 등이다. 하지만 신경망 성과를 향상시키기 위한 네트워크 구조 및 학습요소 최적화 관련한 연구는 기존의 연구들에서 많이 이루어 졌지만 데이터 정규화와 관련한 연구는 아직 많이 이루어지지 않았다. 따라서 본 연구에서는 유전자 알고리즘을 기반으로 하는 정규화 기법을 제시하였다. 최적의 입력데이터 정규화를 위하여 본 연구에서는 우선 각각의 서로 다른 정규화 기법들을 동일 가중치를 두어 일반화 시켰으며 유전자 알고리즘을 이용하여 최적의 가중치를 찾음으로써 최적화된 입력변수 정규화가 이루어지도록 하였다. 제안한 방법론을 검증하기 위하여 부도예측 데이터를 이용하여 실험을 하였으며 제안하는 방법과 기존 다른 방법들간의 비교를 통하여 그 타당성을 검증하였다.. The back-propagation neural network (BPN) has long been successfully applied in bankruptcy prediction problems. Despite its wide application, some major issues must be considered before its use, such as the network topology, learning parameters and normalization methods for the input and output vectors. Previous studies on bankruptcy prediction with BPN have shown that many researchers are interested in how to optimize the network topology and learning parameters to improve the prediction performance. In many cases, however, the benefits of data normalization are often overlooked. In this study, a genetic algorithm (GA)-based normalization transform, which is defined as a linearly weighted combination of several different normalization transforms, will be proposed. GA is used to extract the optimal weight for the generalization. From the results of an experiment, the proposed method was evaluated and compared with other methods to demonstrate the advantage of the proposed method.

GA-optimized Nonlinear Fuzzy Normalization Approach in Back-Propagation Neural Network for Bankruptcy Prediction Modeling

Qiu-yue Tai,Kyung-shik Shin(신경식) 한국지능정보시스템학회 2010 한국지능정보시스템학회 학술대회논문집 Vol.2010 No.11월

The back-propagation neural network (BPN) has been broadly applied to financial distress prediction because their excellent treatment of nonlinear data with learning capabilities. Despite the wide application of BPN, some major issues must be considered before its use, such as the network topologies, learning parameters, and normalization methods for the input and output vectors. Previous studies on bankruptcy prediction with BPN have shown, however, that many researchers are interested in how to optimize the network topologies and learning parameters to improve the network’s prediction performance. In many cases, the benefits of data normalization are overlooked. The most representative method of normalization for BPN is linear scaling, which can reduce the dimensionality of the input space, thus helping speed up the learning phase and improve the classification performance. This method, however, has the limitation of only adjusting the scale of the original data and of not being able to relieve the complicated relationships among the data. An alternative method involves applying the fuzzy set theory to normalize the data for the neural network, because the fuzzy membership function can represent the continuous and complicated values as degree of membership values, and allows the representation of the concepts that can be regarded as falling under more than one category. In this study, a genetic algorithm (GA)-optimized nonlinear fuzzy normalization method was proposed. The polynomial-based nonlinear fuzzy membership function was used to normalize the data within the value of [0, 1]. GA was thus used to find the optimal boundary value of the fuzzy parameters. Based on the results of the experiment that was conducted, the proposed method was evaluated and compared with other methods to demonstrate its advantage.

Electrochemical Properties of Poly(α-methylbenzyl dipropargylamine) Prepared by the Cyclopolymerization of α-Methylbenzyl dipropargylamine

Gui, Tai-Long,Qiu, Jia-Long,Wang, Yue,Shim, Sang-Yeon,Jin, Sung-Ho,Park, Jong-Wook,Lim, Kwon Taek,Gal, Yeong-Soon Informa UK (TaylorFrancis) 2009 Molecular Crystals and Liquid Crystals Vol.498 No.1

Effect of Al2O3 Addition on Microwave Dielectric Properties of BaCo0.194Zn0.116Nb0.69O3 Ceramics

Peijun Liao,Tai Qiu,Jian Yang,Xinyi Lu 대한금속·재료학회 2014 ELECTRONIC MATERIALS LETTERS Vol.10 No.1

This study investigates the effects of Al2O3 addition on sintering, microstructures, phase formation and the microwave dielectric properties of BaCo0.194Zn0.116Nb0.69O3 (BCZN) ceramics prepared by the conventional solidstate route. Addition of low alumina content was found to effectively improve the Q×f value of BCZN ceramics. Addition of an appropriate amount of Al2O3 was found beneficial to the growth and uniformity of the ceramic grain, significantly improving the quality factor of BCZN ceramics. However, the performance of the ceramics deteriorated when an excessive secondary phase, inhomogeneous grain of small size and larger porosity were generated from the BCZN ceramic doped with excessive alumina. An optimum Q × f of 61056 GHz (f = 7.44 GHz) was obtained for the BCZN ceramic doped with 1 wt. % Al2O3 addition sintered at 1350°C for 3 h, which is markedly better than that of pure BCZN ceramic sintered at 1300°C for 3 h (13650 GHz, f = 7.02 GHz).

Biosafety Risk Control Strategies in Laboratory Animal Research

Weng Shun-tai,Li Qu-wen,Gao Ya-dong,Qiu Yu-feng 한국산업안전보건공단 산업안전보건연구원 2024 Safety and health at work Vol.15 No.1

To understand biosafety's current situation in laboratory animal research and risk factors affecting occupational health. Compliance surveys were conducted by questionnaire via Questionnaire Star (an application app on the Internet) in Chinese. Thirty-nine anonymous questionnaires were collected. The surveyed institution has established 24 types of ABSL (Animal Biosafety Laboratory) and biosafety management organizations and systems equipped with safety equipment. Our study also suggests that the principal of the laboratory establishment fails to perform supervision and inspection responsibilities, the inappropriate design of the animal biosafety laboratory, non-standardized personnel training and health management, non-strict waste management, and insufficient emergency management. The administrative department and work units should address certain safety and occupational health risks in laboratory animal research. The author proposes control strategies based on organizational guarantee, personnel management, emergency management, etc., to help prevent risks and ensure occupational health. Due to regional limitations and small sample size, the results may not be generalisable to all parts of the world. However, some of the key common issues may also be present in other regions, so we believe that this research still has some relevance.

Incorporation of Titanium into H-ZSM-5 Zeolite via Chemical Vapor Deposition: Effect of Steam Treatment

Xu, Cheng-Hua,Jin, Tai-Huan,Jhung, Sung-Hwa,Hwang, Jin-Soo,Chang, Jong-San,Qiu, Fa-Li,Park, Sang-Eon Korean Chemical Society 2004 Bulletin of the Korean Chemical Society Vol.25 No.5

Ti-ZSM-5 prepared by secondary synthesis, from the reaction of H-ZSM-5 with vapor phase $TiCl_4$, was characterized with several physicochemical techniques including FT-IR and UV/VIS-DRS. It was found that zeolite structure, surface area and pore volume did not change, and the framework aluminum could not be replaced by titanium atom during the secondary synthesis of Ti-ZSM-5. The incorporation of titanium into the framework might be due to reaction of $TiCl_4$with the silanol groups associated with defects or surface sites. The formation of extra-framework titanium could not be avoided, unless the samples were further treated by water vapor at 550 $^{\circ}C$ or higher temperature. High temperature steam treatment of Ti-ZSM-5 prepared by chemical vapor deposition with $TiCl_4$was efficient to prevent the formation of non-framework titanium species. Ti-ZSM-5 zeolites prepared in this work contained only framework titanium species and exhibited improved catalytic property close to TS-1 prepared by hydrothermal synthesis.

Carbon-supported Ni nanoparticles for efficient CO ₂ electroreduction

Jia, Mingwen,Choi, Changhyeok,Wu, Tai-Sing,Ma, Chen,Kang, Peng,Tao, Hengcong,Fan, Qun,Hong, Song,Liu, Shizhen,Soo, Yun-Liang,Jung, Yousung,Qiu, Jieshan,Sun, Zhenyu Royal Society of Chemistry 2018 Chemical Science Vol.9 No.47

<▼1><P>Carbon-coated Ni nanoparticles supported on N-doped carbon enable efficient electroreduction of CO<SUB>2</SUB> to CO comparable to single Ni sites.</P></▼1><▼2><P>The development of highly selective, low cost, and energy-efficient electrocatalysts is crucial for CO<SUB>2</SUB> electrocatalysis to mitigate energy shortages and to lower the global carbon footprint. Herein, we first report that carbon-coated Ni nanoparticles supported on N-doped carbon enable efficient electroreduction of CO<SUB>2</SUB> to CO. In contrast to most previously reported Ni metal catalysts that resulted in severe hydrogen evolution during CO<SUB>2</SUB> conversion, the Ni particle catalyst here presents an unprecedented CO faradaic efficiency of approximately 94% at an overpotential of 0.59 V, even comparable to that of the best single Ni sites. The catalyst also affords a high CO partial current density and a large CO turnover frequency, reaching 22.7 mA cm<SUP>–2</SUP> and 697 h<SUP>–1</SUP> at –1.1 V (<I>versus</I> the reversible hydrogen electrode), respectively. Experiments combined with density functional theory calculations showed that the carbon layer coated on Ni and N-dopants in carbon material both play important roles in improving catalytic activity for electrochemical CO<SUB>2</SUB> reduction to CO by stabilizing *COOH without affecting the easy *CO desorption ability of the catalyst.</P></▼2>

Ultrasound-enhanced Subcritical Water Extraction of Naphthoquinone Pigments from Purple Gromwell (Lithospermum erythrorhizon) to Higher Yield and Bioactivity

Ri-fu Yang,Ping-ping Huang,Tai-qiu Qiu 한국식품과학회 2013 Food Science and Biotechnology Vol.22 No.3

Naphthoquinone pigments were extracted from purple gromwell (Lithospermum erythrorhizon) via ultrasoundenhanced subcritical water extraction (USWE). USWE was compared with supercritical CO2 fluid extraction (SCE) and Soxhlet extraction (SE) on the extraction yield of naphthoquinone pigments and the components and various activities of the extracts. The results show that the extraction yield of naphthoquinone pigments by USWE was the highest, up to 0.31%, while the yields achieved by SE and SCE were 0.21 and 0.19%, respectively. The 6major compounds were detected in these 3 extract samples but the USWE sample contained the fewest impurities. The extracts obtained with USWE showed the best inhibitory effects on Escherichia coli and Staphylococcus aureus,followed by that with SE. However, the extracts obtained with SE showed the highest DPPH scavenging capacity and reducing action.

A Study on Resistance to Blast in japonica Rice Breeding

Cheng Huan Cui,De Tang Zou,Rong Tian Li,Tai Quan Qiu 한국육종학회 1997 한국육종학회 학술발표회 발표요지 Vol.29 No.2

The influence factors on CeSn0.8W0.6Ox/TiO2 for catalytic removals of NO, CO and C3H8

Guorong Sui,Zhiwei Xue,Dan Zhou,Yan Wang,Yuesong Shen,Yuhao Zong,Youlin Liu,Tai Qiu,Shemin Zhu 한국공업화학회 2017 Journal of Industrial and Engineering Chemistry Vol.51 No.-

Series of CeSn0.8W0.6Ox/TiO2 catalysts were tested for selective catalytic reduction of NO and forsynergistic catalytic removals of CO and C3H8 from diesel engine exhaust. Results revealed that catalyst12%-CeSn0.8W0.6Ox/TiO2 calcined at 500 C exhibited the optimal catalytic performance for NH3-SCR ofNO. The catalyst obtained more than 90% NO conversion at a wide temperature range of 252–456 C. BothCO and C3H8 could be oxidized into CO2 by the optimized catalyst. Moreover, excellent redox property,rich surface acidity and big specific surface area were the promotional factors for good catalyticperformance in catalytic removals of NO, CO and C3H8.