應用資料探勘於全額交割股恢復交易之探討

Using data mining approach for prediction of resuming stocks requiring full delivery to normal trades

陳琇宜(Show-Yi Chen)

指導教授 ： 胡念祖

國立虎尾科技大學/管理學院/資訊管理研究所在職專班/碩士(2009年)

摘要

2008年以來發生全球金融風暴至今，全世界許多企業面臨財務危機的窘境甚至面臨倒閉，加上近幾年經濟不景氣、原物料的上漲等因素之影響，使得很多危機造成國內企業因週轉不靈、跳票或申請重整等原因，被打入全額交割。然而並不是每間危機公司都能夠順利的從失敗中找到方法，重新找到企業的重生契機，究竟影響企業能否順利恢復正常交易的主要因素為何，實值得探討。 本研究住要目的乃是欲探勘出過去在台灣股市中，若干因財務危機而被打入全額交割的公司能成功恢復（櫃）的因素。 在此本文我們應用以決策樹為基礎之資料探勘方法來探索出因財務危機而被打入全額交割股的上市、上櫃公司是否能成功重新上市上櫃成功的分類法則。更進一步，利用以決策樹演算法為基礎之boosting ensemble 方法所建立之多重分類器模型被建立。由實驗數據顯示利用所建立之多重分類器模型使得分類準確率能成功被提升，且型二錯誤也能成功的被減小。此外、所探勘出之法則可以被發展成為判斷因財務危機之全額交割股的公司是否能成功重新上市上櫃之電腦決策模型如同建立專家系統一般。

關鍵字

全額交割 ； 資料探勘 ； 分類法則

並列摘要

In recent years, a number of financial crises have made prediction of resuming stocks requiring full delivery to normal trades to become a noticeable topic to both practices and academy. In order to make their decisions correctly in time, all of the creditors, analysts, investors and regulators wish to predict whether financially distressed firms will be able to emerge based on the information available at the time of the company’s stocks requiring full delivery. However, evaluating the feasibility of financial reorganization success is complex. In this research, we employed decision tree-based mining techniques to develop a prediction model. Besides, the multi-learner model constructed by boosting ensemble approach with decision tree algorithm is used to enhance the prediction accuracy rate. The empirical results show that the classification accuracy has been improved by using multi-leaner model in terms of less Type II errors. In particular, the extracted rules from the data mining approach can be developed as a computer model for the prediction and like expert systems.

並列關鍵字

Stocks requiring full delivery ； Data mining ； Classification rule

參考文獻

[1] Altman, E. (1968), “Financial ratios, discriminant analysis and the prediction of corporate bankruptcy,” Journal of Finance, 23, pp. 589-609.

[2] Breiman, L (1996), “Bagging predictors,” Machine Learning, 24 (2), pp.123-140.

[3] Berry, A. J. M. and Limoff, Gordon. (1997), Data Mining Techniques, John Wiley & Sons, Canada.

[4] Becchetti, L., & Sierra, J. (2003), “Bankruptcy risk and productive efficiency in manufacturing firms,” Journal of Banking and Finance, 27(11), pp. 2099-2120.

[5] Berlin, M. (1996), “For Better and for Worse: Three Lending Relationships,” Business Review, pp.3-12.