A Study of the Financial Crisis of the Credit Department of Farmer Association -- the Genetic Fuzzy Neural Network Approach
碩士 === 東吳大學 === 經濟學系 === 92 === In the past few years, the financial crisis of the credit department of farmer association occurs from time to time. Other than the general social environment change and the insufficiency of the staff quality, there are defects in common, such as the insolvency,low ne...
Main Authors: | , |
---|---|
Other Authors: | |
Format: | Others |
Language: | zh-TW |
Published: |
2004
|
Online Access: | http://ndltd.ncl.edu.tw/handle/99312628454106365490 |
id |
ndltd-TW-092SCU00389042 |
---|---|
record_format |
oai_dc |
spelling |
ndltd-TW-092SCU003890422016-06-15T04:17:27Z http://ndltd.ncl.edu.tw/handle/99312628454106365490 A Study of the Financial Crisis of the Credit Department of Farmer Association -- the Genetic Fuzzy Neural Network Approach 應用遺傳演化模糊類神經網路對農會信用部財務危機預警之研究 FAN.YING-TE 范英德 碩士 東吳大學 經濟學系 92 In the past few years, the financial crisis of the credit department of farmer association occurs from time to time. Other than the general social environment change and the insufficiency of the staff quality, there are defects in common, such as the insolvency,low net worth, and the poor earnings. The management morality, the limitation of the promotion of the credit department of farmer association, and the over-conservative business policy are the other factors for these phenomena. All of these will be shown on the financial reports, except for the human-factor. Therefore this paper will explore the CAMELS theory both in local and international essays, and use it as the base of the parameters selection. In addition to the traditional econometrics techniques, artificial intelligence will be formed the financial crisis alarm models. Of course we hope to find the most effective one to provide precaution for all parties concerned in our economy, and to minimize the impact by early reaction. In our empirical study, we use the 285 credit department of farmer associations in 2000 as a sample. But the financial crisis occurred in 26 of them, and all were forced take-over by our superior authority. Therefore, we will take the former sample as the problem credit iii department, and the others as the normal department. As the related studies have shown the data of the one year before provides the most convincing estimation. So we will take this data as the sample only. For the Pareto,s distribution ,80-20 rate will be taken in a random way as the training and testing periods. That is 57 observations will be used as the testing sample. In addition, we use four financial crisis models--the Logit Regression, Back Propagation Networks (BPN), Genetic Algorithms Back Propagation Network(GABPN), and Fuzzy Genetic Algorithms Back Propagation Network(FGABPN)--to predict the bankruptcy. Furthermore, the Wilcoxon RANK test will be applied to confirm the performance of these four models, i.e. to find the best forecast model. The empirical results show that the superiority of Artificial Intelligence as bellow, A. On the accuracy, it goes as GAPBN > FGAPBN > Logit >BPN. B. Upon the square error comparison, FGABPN and GAPBN show not so much difference, but FGABPN is obviously better than BPN and Logit. Lin Wei-Yuan 林維垣 2004 學位論文 ; thesis 106 zh-TW |
collection |
NDLTD |
language |
zh-TW |
format |
Others
|
sources |
NDLTD |
description |
碩士 === 東吳大學 === 經濟學系 === 92 === In the past few years, the financial crisis of the credit department of farmer association occurs from time to time. Other than the general social environment change and the insufficiency of the staff quality, there are defects in common, such as the insolvency,low net worth, and the poor earnings. The management morality, the limitation of the promotion of the credit department of farmer association, and the over-conservative business policy are the other factors for these phenomena. All of these will be shown on the financial reports, except for the human-factor. Therefore this paper will explore the CAMELS theory both in local and international essays, and use it as the base of the parameters selection. In addition to the traditional econometrics techniques, artificial intelligence will be formed the financial crisis alarm models. Of course we hope to find the most effective one to provide precaution for all parties concerned in our economy, and to minimize the impact by early reaction. In our empirical study, we use the 285 credit department of farmer associations in 2000 as a sample. But the financial crisis occurred in 26 of them, and all were forced take-over by our superior authority. Therefore, we will take the former sample as the problem credit iii
department, and the others as the normal department. As the related studies have shown the data of the one year before provides the most convincing estimation. So we will take this data as the sample only. For the Pareto,s distribution ,80-20 rate will be taken in a random way as the training and testing periods. That is 57 observations will be used as the testing sample. In addition, we use four financial crisis models--the Logit Regression, Back Propagation Networks (BPN), Genetic Algorithms Back Propagation Network(GABPN), and Fuzzy Genetic Algorithms Back Propagation Network(FGABPN)--to predict the bankruptcy. Furthermore, the Wilcoxon RANK test will be applied to confirm the performance of these four models, i.e. to find the best forecast model. The empirical results show that the superiority of Artificial Intelligence as bellow, A. On the accuracy, it goes as GAPBN > FGAPBN > Logit >BPN. B. Upon the square error comparison, FGABPN and GAPBN show not so much difference, but FGABPN is obviously better than BPN and Logit.
|
author2 |
Lin Wei-Yuan |
author_facet |
Lin Wei-Yuan FAN.YING-TE 范英德 |
author |
FAN.YING-TE 范英德 |
spellingShingle |
FAN.YING-TE 范英德 A Study of the Financial Crisis of the Credit Department of Farmer Association -- the Genetic Fuzzy Neural Network Approach |
author_sort |
FAN.YING-TE |
title |
A Study of the Financial Crisis of the Credit Department of Farmer Association -- the Genetic Fuzzy Neural Network Approach |
title_short |
A Study of the Financial Crisis of the Credit Department of Farmer Association -- the Genetic Fuzzy Neural Network Approach |
title_full |
A Study of the Financial Crisis of the Credit Department of Farmer Association -- the Genetic Fuzzy Neural Network Approach |
title_fullStr |
A Study of the Financial Crisis of the Credit Department of Farmer Association -- the Genetic Fuzzy Neural Network Approach |
title_full_unstemmed |
A Study of the Financial Crisis of the Credit Department of Farmer Association -- the Genetic Fuzzy Neural Network Approach |
title_sort |
study of the financial crisis of the credit department of farmer association -- the genetic fuzzy neural network approach |
publishDate |
2004 |
url |
http://ndltd.ncl.edu.tw/handle/99312628454106365490 |
work_keys_str_mv |
AT fanyingte astudyofthefinancialcrisisofthecreditdepartmentoffarmerassociationthegeneticfuzzyneuralnetworkapproach AT fànyīngdé astudyofthefinancialcrisisofthecreditdepartmentoffarmerassociationthegeneticfuzzyneuralnetworkapproach AT fanyingte yīngyòngyíchuányǎnhuàmóhúlèishénjīngwǎnglùduìnónghuìxìnyòngbùcáiwùwēijīyùjǐngzhīyánjiū AT fànyīngdé yīngyòngyíchuányǎnhuàmóhúlèishénjīngwǎnglùduìnónghuìxìnyòngbùcáiwùwēijīyùjǐngzhīyánjiū AT fanyingte studyofthefinancialcrisisofthecreditdepartmentoffarmerassociationthegeneticfuzzyneuralnetworkapproach AT fànyīngdé studyofthefinancialcrisisofthecreditdepartmentoffarmerassociationthegeneticfuzzyneuralnetworkapproach |
_version_ |
1718305593720045568 |