Title

財務預警混合模式之特徵篩選與模型建構-以台灣電子產業為例

Translated Titles

The Feature Selection and Hybrid Model Construction of Bankruptcy Prediction on Taiwan Electronic Industry

DOI

10.6841/NTUT.2011.00631

Authors

劉均猷

Key Words

財務預警 ; 破產預測 ; 逐步羅吉斯回歸 ; 倒傳遞類神經 ; 一般迴歸類神經 ; Financial distress prediction ; Feature Selection ; Stepwise Logistic Regression ; Back-Propagation Neural Network ; General Regression Neural Network

PublicationName

臺北科技大學工業工程與管理系碩士班學位論文

Volume or Term/Year and Month of Publication

2011年

Academic Degree Category

碩士

Advisor

羅淑娟

Content Language

繁體中文

Chinese Abstract

財務預警模型有助於投資者、政府或企業藉由模型對於公司經營狀況作一整體性預測與研判,財務預警模式相關文獻多以追求分類預測的準確度為研究目標,對於建構模型的變數或選擇變數方法的效果較少著墨,類神經分類模型在財務預警模型有不錯的分類預測效果,但類神經模型本身並無變數挑選機制,本研究以特徵挑選與分類預測兩階段去建構財務預警混合模型,特徵變數挑選有關聯係數(Eta square)和逐步羅吉斯回歸分析(Stepwise Logistic Regression)兩種方法,分類預測方法有倒傳遞類神經網路(Back-Propagation Neural Network,BPNN)與一般迴歸類神經網路(General Regression Neural Network,GRNN)兩種類神經網路方法。利用以上兩階段各兩種方法做混合模型,共4種混合模型(Stepwise LR+BPNN、Stepwise LR+GRNN、Eta square +BPNN、Eta square +GRNN),探討混合模型對公司危機預測的績效。另外利用Stepwise LR、BPNN、GRNN去做3種單一模型,探討不同的混合與單一模型對公司破產預測績效的影響,本研究採用預測準確率及檢定力(危機公司預測為危機的能力)作為績效定義,除找出績效最優的財務預警模型外,也討論變數挑選階段對預警模型的貢獻和有顯著貢獻的特徵變數,研究資料以台灣上市電子公司1999年至2006年的季資料作為實驗樣本,實驗結果顯示,兩階段類神經模型都較單一類神經模型在測試的平均準確率高出接近6%,而特徵篩選方法與類神經方法間也存有交互作用,以BPNN搭配Stepwise LR和GRNN搭配Eta square有較佳績效。

English Abstract

A financial prediction model can help investors, government and cooperation executioners to make an overall prediction or judgment on the operation status of a firm. The most related literatures of prediction model pursue the high prediction correctness as their research targets. They rarely discussed the impact of feature selection on prediction models. The method of neural network has high classification correctness in the financial prediction model, but it has no direct mechanism on the feature selection. In this research, we use a two-step concept to construct a hybrid model with feature selection step and classification step. The feature selection step has two experimental methods, Eta square and Stepwise logistic regression. And the classification step chooses the well known Back-propagation neural network and the General regression neural network. There are four hybrid models, Stepwise LR+BPNN, Stepwise LR+GRNN, Eta square +BPNN and Eta square +GRNN, discussed in this research. The stepwise LR, BPNN and GRNN can also be prediction models. Therefore, we discuss the bankrupt prediction performance of these seven models. The definition of performance includes the overall prediction correctness and power which is the ability to predict bankruptcy for those bankrupt firms. The contribution of feature selection step and important bankrupt features are also discussed in this study. Our experimental data are from 1999 to 2006 Taiwan listed electronic companies. Experimental results show that hybrid models are superior to single neural networks about 6% in the correctness. There are interactions between neural networks and statistical methods. The better performances appear in Stepwise LR with BPNN and Eta square with GRNN.

Topic Category 管理學院 > 工業工程與管理系碩士班
工程學 > 工程學總論
社會科學 > 管理學
Reference
  1. [2] 許溪南、歐陽豪、陳慶芳,盈餘管理、公司治理與財務預警模型之建構,金融風險管理季刊,第3 卷,第3 期,2007,第1-40 頁。
    連結:
  2. [3] 羅淑娟,林晶璟,陳義方,應用邏吉斯迴歸技術探討財務危機預警變數與資料長度之適用性研究--以台灣上市電子產業為例,台北科技大學學報,四十二之二期,2009。
    連結:
  3. [4] C. M. Daily, and D. R. Dalton, "Bankruptcy and Corporate Governance: The Impact of Board Composition and Structure, " Academy of Management, vol.37, 1994, pp. 1603-1617.
    連結:
  4. [5] D. F. Specht, "A General Regression Neural Network , "IEEE Transactions on Neural Network, vol.26, 1991 , pp.568-576.
    連結:
  5. [6] E. B. Deakin, "A discriminate analysis of predictors of business failure , " Journal of Accounting Research ,vol.Spring , 1972,pp.167-179.
    連結:
  6. [7] E. I. Altman, "Financial ratios—discriminant analysis and the prediction of corporate bankruptcy using capital market data, "Journal of Finance , vol.23 No. 4 , 1968, pp.589-609.
    連結:
  7. [8] E. I. Altman, G. Marco and F. Varetto , " Corporate Distress Diagnosis: Comparisons Using Linear Discriminant Analysis and Neural Networks, "Journal of Banking and Finance, vol. 18, 1994, pp. 505-529.
    連結:
  8. [9] F. A. Amir., "Bankruptcy Prediction for Credit Risk Using Neural Networks :A Survey and New Results, " IEEE Transaction JNL on Neural Networks, vol.12, No. 4, 2001, pp. 929-935.
    連結:
  9. [11] G. Zhang , M. Y. Hu , B. E. Patuwo and D. C. Indro, "Artificial neural networks in bankruptcy prediction :General framework and cross-validation analysis, "European Journal of Operational Research ,vol.116,1999, pp.16-32.
    連結:
  10. [12] J. A. Ohlson, "Financial ratios and the probability prediction of bankruptcy, " Journal of Accounting Research, vol.18 , No. 1 , 1980 , pp.109-131.
    連結:
  11. [13] J. H. Mina and Y. C. Lee, "Bankruptcy prediction using support vector machine with optimal choice of kernel function parameters," Expert Systems with Applications, vol.28, 2005, pp.603-614.
    連結:
  12. [14] K. S. Shin , T. S. Lee and H .J. Kim, "An application of support vector machines in bankruptcy prediction model, " Expert Systems with Applications, vol.28, Issue 1, 2005,pp. 127-135.
    連結:
  13. [15] M. E. Zmijewski, "Methodological issues related to the estimation of financial distress prediction models," Journal of Accounting Research, vol. 22 , 1984, pp. 59–82.
    連結:
  14. [16] M. Odom and R. Sharda, "A neural network model for bankruptcy prediction, "
    連結:
  15. [17] P. Jardin, "Predicting bankruptcy using neural networks and other classification methods: The influence of variable selection techniques on model accuracy, "Neurocomputing, vol.73, 2010, pp.2047-2060.
    連結:
  16. [18] P. Ravisankar, V. Ravi and I. Bose, "Failure prediction of dotcom companies using neural network–genetic programming hybrids," Information Sciences, vol. 180 , 2010, pp.1257–1267.
    連結:
  17. [19] S. Balcaen and H. Ooghe, "35 years of studies on business failure: an overview of the classic statistical methodologies and their related problems, "The British Accounting Review, vol.38, 2006, pp. 63-93.
    連結:
  18. [20] S. L. Lin, "A new two-stage hybrid approach of credit risk in banking industry, " Expert Systems with Applications, vol.36 , 2009, pp.8333-8341.
    連結:
  19. [21] S. Wu and X. Lu, "A model on predicting the bankruptcy of Chinese public company," Economic Research, vol.6 (8) , 2001, pp. 46-55.
    連結:
  20. [22] T. H. Lin, "A cross model study of corporate financial distress prediction in Taiwan: Multiple discriminant analysis, logit, probit and neural networks models, " Neurocomputing, vol.72, 2009, pp.3507-3516.
    連結:
  21. [23] T. S. Lee and I. F. Chen , "A two-stage hybrid credit scoring model using artificial neural networks and multivariate adaptive regression splines, "Expert Systems with Application, vol.28 , No. 4 , 2005, pp.743-752.
    連結:
  22. [24] W. H. Beaver, "Financial ratios and predictors of failure, "Journal of Accounting Research: Empirical Research in Accounting—Selected Studies, vol.4 , 1966, pp.71-111.
    連結:
  23. [25] W. Hopwood, J.C. McKeown and J.F. Mutchler, "A reexamination of auditor versus model accuracy within the Context of the Going-Concern Opinion Decision, " Contemporary Accounting Research, vol.10, No 2 , 1994, pp. 409-431.
    連結:
  24. [26] 洪立劼,應用倒傳遞類神經與序列探勘技術建構企業財務危機預警模型─以台灣電子產業為例,碩士論文,國立台北科技大學工業工程與管理學系研究所,台北,2004。
    連結:
  25. [27] 陳義方,應用邏吉斯迴歸技術探討財務危機預警變數與資料長度之適用性研究-以台灣電子產業為例,碩士論文,國立台北科技大學工業工程與管理學系研究所,台北,2009。
    連結:
  26. [29] 蔡明賢,用一般迴歸類神經網路與序列探勘技術建構企業財務危機預警模型-以台灣電子產業為例,國立台北科技大學工業工程與管理學系研究所,台北,2008。
    連結:
  27. [30] D. E. Rumelhart , J. L. McClelland ,Parallel distributed processing: exploration in the cognition. , MA, MIT, Press Cambridge,1986.
    連結:
  28. 參考文獻
  29. 書籍
  30. [1] 江建良,統計學,台北:高立出版公司,2010。
  31. 期刊論文
  32. [10] Fletcher , Desmond, Goss and Ernie, "Forecasting with Neural Networks:An Application Using Bankrupt Data, " Information & Management, vol. 24,1993, pp.159-67.
  33. The IEEE Internationa Joint Conference on Neural Networks, vol.1 , 1990, pp.163–168.
  34. 學位論文
  35. [28] 曾冠人,應用一般迴歸神經網路法構建財務危機預警模式,碩士論文,國立交通大學工業工程與管理學系研究所,新竹,2004。