Erkan Cetiner

Classifier Performances for Credit Risk Analysis: a Hybrid Classification Approach on Credit Risk Analysis Erkan Cetiner

This work is prepared for a Master Research Thesis. The main objective of the work is gathering single classification techniques together as one unique hybrid classifier. Experiments made on different data-sets and results are compared in terms of accuracy and precision. Logistic regression, support vector machines, artificial neural networks and naive bayes approach are examined throughout the research. A hybrid model based on average weighting mechanism developed by using those single classifiers.