Institutional Repository [SANDBOX]
Technical University of Crete
EN  |  EL

Search

Browse

My Space

Analytical methodologies for business failure prediction in the energy sector. A comparative analysis for European Companies

Iliaki Georgia

Full record


URI: http://purl.tuc.gr/dl/dias/E284D4D6-BB2C-47AA-A4BA-09C20C424B60
Year 2018
Type of Item Diploma Work
License
Details
Bibliographic Citation Georgia Iliaki, "Analytical methodologies for business failure prediction in the energy sector. A comparative analysis for European Companies", Diploma Work, School of Production Engineering and Management, Technical University of Crete, Chania, Greece, 2018 https://doi.org/10.26233/heallink.tuc.78962
Appears in Collections

Summary

This study references the European energy sector and the plans for a united European energy sector but its primary purpose is to extend the paper of Doumpos et al (2017) and investigates different analytical methodologies for financial distress prediction in a very large sample from different European countries and energy sectors. The data spans between the period of 2012-2016. The methodologies used are discriminant analysis (linear and quadratic), logistic regression, classification trees (simple, boosted and bagged) and machine learning methods (neural networks and support vector machines). All the methods were applied on the whole sample at first and after on specific energy sectors and countries. Whole sample analysis proved more effective than country and sector analysis and the most accurate method was the bagged trees. For the sector and country analysis the most effective method was the neural networks.

Available Files

Services

Statistics