We found a match
Your institution may have access to this item. Find your institution then sign in to continue.
- Title
HALKA AÇIK FİNANS DIŞI ŞİRKETLERDE SÜREKLİLİK RİSKİNİN KARAR AĞACI MODELİ İLE ÖNGÖRÜLMESİ.
- Authors
ÜNKAYA, Gülümser; SAYIN, Gürkan
- Abstract
This research aims to determine the potential of decision trees, a subclass of machine learning models, in going concern prediction among non-finance public companies. Data used in the study to construct the model includes nonfinance Turkish companies traded publicly in 1999-2016 period. R statistical language and related machine learning libraries (rpart and partykit) are utilized to train the model. Performance of the model is estimated via cross-validation, a commonly accepted method. The study shows that, the decision tree model applied to the data attains a prediction performance of 91%. Besides being very effective machine learning models for the current scenario, decision trees also appear to be very easy to comprehend and build. It is foreseen that, decision tree models can be utilized by credit analysts at any level and be incorporated to any credit analysis software without much effort.
- Subjects
DECISION trees; CREDIT analysis; PROGRAMMING languages; MACHINE learning; PUBLIC companies
- Publication
Financial Analysis / Mali Cozum Dergisi, 2019, Vol 29, Issue 156, p13
- ISSN
1303-5444
- Publication type
Article