Rating Companies with Support Vector Machines

Arbeitspapier

The goal of this work is to introduce one of the most successful among recently developed statistical techniques - the support vector machine (SVM) - to the field of corporate bankruptcy analysis. The main emphasis is done on implementing SVMs for analysing predictors in the form of financial ratios. A method is proposed of adapting SVMs to default probability estimation. A survey of practically and commercially applied methods is given. This work proves that support vector machines are capable of extracting useful information from financial data although extensive data sets are required in order to fully utilise their classification power.

Show object

• Language: Englisch

References and Relationships

• Bibliographic citation: Series: DIW Discussion Papers ; No. 416

Classification and Topics

• Classification: Wirtschaft ; Neural Networks and Related Topics ; Bankruptcy; Liquidation ; Semiparametric and Nonparametric Methods: General
• Subject: Support vector machines ; Company rating ; Default probability estimation ; Kreditwürdigkeit ; Mustererkennung ; Schätzung ; Theorie ; Vereinigte Staaten ; support vector machine

Contributors, Places and Time

• Event: Geistige Schöpfung
• (who): Schäfer, Dirk ; Moro, R. A. ; Härdle, Wolfgang Karl
• Event: Veröffentlichung
• (who): Deutsches Institut für Wirtschaftsforschung (DIW)
• (where): Berlin
• (when): 2004

Further information

• Handle: http://hdl.handle.net/10419/18111