Reflections on Financial Distress Prediction Model Using Support Vector Machine and Logit Model
碩士 === 國立高雄第一科技大學 === 財務管理所 === 92 === This paper examines three issues-choice-based sample bias, predictor selecting, and consistency- in financial distress prediction model. Support Vector Machine and Logit regression are proposed to predict financial distress. The selected techniques are Altman...
| Main Authors: | , |
|---|---|
| Other Authors: | |
| Format: | Others |
| Language: | zh-TW |
| Published: |
2004
|
| Online Access: | http://ndltd.ncl.edu.tw/handle/93976231169450678500 |
| Summary: | 碩士 === 國立高雄第一科技大學 === 財務管理所 === 92 === This paper examines three issues-choice-based sample bias, predictor selecting, and consistency- in financial distress prediction model. Support Vector Machine and Logit regression are proposed to predict financial distress. The selected techniques are Altman analysis and forward Wald selection methods. In empirical results, the most important determinants of financial distress are leverage ratio and operating profit ratio. The choice-based sample bias is statistically significant. The predict results doesn’t decrease following time passing. The model with support vector machine and forward Wald selection method produce superior results for predicting bankrupt companies.
|
|---|