Translated Titles




Key Words

財務危機 ; 預測模型 ; 成長式財務變數擷取方法 ; 灰色適應共振網路 ; financial distress ; GreyART network ; growing extraction method for financial variable ; prediction model



Volume or Term/Year and Month of Publication


Academic Degree Category




Content Language


Chinese Abstract

本研究論文所採用之財務危機預測模型是以灰色適應共振網路(GreyART network)為核心,而所使用的數據是從台灣經濟新報資料庫挑選出來的。文中,灰色適應共振網路之每一訓練(training)或測試(test)範例均是由18個不同財務比例所構成的,其中訓練範例為涵蓋民國89至91年之資料,而測試範例僅包括民國92年之資料。首先,為使所提財務危機預測模型能獲得最佳結果,本文將分割品質(partition quality)和歸類精確度(classification accuracy)之乘積定義為績效指標(performance index),具最大績效指標所對應之結果即所謂之最佳結果,以此方式所獲得之訓練與測試最佳結果分別為88.24%和61.29%。再者,本文也提出一成長式財務變數擷取方法(growing extraction method for financial variables)期能同時達到以較少的財務變數來建立財務危機預測模型和提高歸類正確度之目的。模擬結果可知,以股東權益成長率(stockholder’s equity growth rate)、總資產週轉成長率(total assets turnover growth rate)、稅前淨利率(earning before taxes rate)、營收成長率(revenue growth rate)、淨值/總資產(stockholder / total assets)、每股淨值(EPS)、稅前淨利成長率(earning before taxes growth rate)和負債比率(debt ratio)等8個所擷取之財務比率變數所建立的模型其結果最好,訓練與測試結果分別提升至94.12%和93.55%,且在此條件之下,所提方法僅產生4個聚類(clusters),1個屬於財務正常公司、3個屬於財務危機公司。此外,以部分所擷取財務變數所建立的模型,不論在訓練或測試階段,大部分之歸類正確度均可高於90%,也就是說,所提之變數擷取方法的確能改善歸類正確度。

English Abstract

This study attempts to use the GreyART network to construct a financial distress prediction model. The inputs applied to the GreyART network are the historical data containing 18 different financial ratios. They are collected from the Taiwan Economic Journal Data Bank covering the period 2000-03. In order to determine the best result the proposed method can attain, a new performance index defined as the product of the partition quality and the classification accuracy is proposed to solve this problem. The best results achieve the classification hit rates of 88.24% and 61.29% for the training and test phases, respectively, where the test result represents the case of one-year-ahead prediction. This study also proposes a growing extraction method for financial variables not only to further improve the classification ability in the training and test phases, but also to use fewer extracted variables to build the financial distress prediction model. Simulation results show that the best result is the one used by the following 8 extracted variables -- stockholder’s equity growth rate, total assets turnover growth rate, earning before taxes rate, revenue growth rate, stockholder / total assets, EPS, earning before taxes growth rate, and debt ratio. The corresponding classification hit rates of the training and test phases are rise to 94.12% and 93.55%, respectively. While using those 8 extracted variables, there are only four clusters, 1 for healthy class and 3 for distressed class, to be generated. Additionally, most of cases that use some extracted variables can attain the classification accuracy above 90% in the training and test phases. That is to say, the classification ability is indeed improved with the proposed variable extraction method.

Topic Category 經營設計學院 > 事業經營學系所
社會科學 > 管理學
  1. Altman, E. I. “Financial ratios, discriminant analysis and the prediction of corporate bankruptcy.” Journal of Finance 23, no. 4 (1968): 589-609.
  2. Beaver, W. H. “Financial ratios as predictors of failure.” Empirical Research in Accounting: Selected Studies 5 (1966): 71-111.
  3. Calderon, T. G., and J. J. Chen. “A roadmap for future neural networks research in auditing and risk assessment.” International Journal of Accounting Information Systems 3 (2002): 203-226.
  4. Coats, P. K., and L. F. Fant. “Recognizing financial distress patterns using a neural network tool.” Financial Management 22, no. 3 (1993): 142-155.
  5. Grossberg, S. “Adaptive pattern classification and universal recording: I. Parallel development and coding of neural feature detectors.” Biological Cybernetics 23 (1976): 121-134.
  6. . “Adaptive pattern classification and universal recording: II. Feedback, expectation, olfaction, and illustrations.” Biological Cybernetics 23 (1976): 187-202.
  7. Hsia, K. H., and J. H. Wu, “A study on the data preprocessing in grey relational analysis.” Journal of Chinese Grey System 1, no. 1 (1998): 47-54.
  8. Ohlson, J. A. “Financial ratios and the probabilistic prediction of bankruptcy.” Journal of Accounting Research 18, no. 1 (1980): 109-131.
  9. Vellido, A., P. J. G. Lisboa, and J. Vaughan. “Neural networks in business: a survey of applications (1992-1998).” Expert Systems with Applications 17 (1999) 51-70.
  10. Wong, B. K., T. A. Bodnovich, and Y. Selvi. “Neural network applications in business: a review and analysis of the literature (1988-95).” Decision Support System 19 (1997): 301-320.
  11. Yeh, M. F., and S. S. Chiang, “GreyART network for data clustering.” Neurocomputing 67 (2005): 313-320.
  12. 陳肇榮,運用財務比率預測企業財務危機之實證研究,博士論文,國立政治大學財政研究所,民72。
  13. 郭瓊宜,類神經網路在財務危機預警模式之應用,碩士論文,淡江大學管理科學研究所,民83。
  14. 黃瓊華,現金流量與傳統財務比率預測企業失敗之研究,碩士論文,中華大學工業研究所,民84。
  15. 趙勃軒,以財務比率分析區別台灣地區上市電子公司經營績效,碩士論文,大同大學事業經營研究所在職專班,民91。
  16. 蔡松林,灰色理論於公司財務危機分析,碩士論文國立台灣科技大學資訊工程系,民93。
  17. Hamerly, G., and C. Elkan, “Learning the k in k-means.” Proceeding of 17th Annual Conference on Neural Information Processing Systems, British Columbia, Canada, 2003.
  18. Odom, M. D. and R. Sharda. “A neural network model for bankruptcy prediction.” Proceedings of 1990 International Joint Conference on Neural Networks, San Diego, CA, vol. 2, pp. 163-168, 1990.