Credit evaluation model using naïve bayes classifier a case of a Kenyan Commercial Bank
Abstract
With the increasing demand for credit facilities for the purpose of development, more and more financial institutions
are being established to cater for the need. Acquiring these facilities from the institutions sometimes prove slow and
inefficient due to the model adopted for credit evaluation. Reliance on traditional methods for instance, a checklist
of bank rules, conventional statistical methods and personal judgment in evaluating credit worthiness makes the
process slow and such judgments could be biased. Effective models are required to help mitigate these day-to-day
challenges. This study examines the relevance of Naïve Bayes Classifier as an enabling tool in credit decision that
can automatically evaluate credit applications based on customer‟s biographic, demographic and behavioural
characteristics. Data used is obtained from one of the commercial banks in Kenya. Feature selection is performed on
the data in order to eliminate redundant and less relevant variables. A model using Naïve Bayes Classifier algorithm
is developed and its classification performance evaluated. Results show that Naïve Bayes Classifier can be used as a
credit decision tool that can speed up and improve efficiency of the process. It also shows that using significant
variables improves the model‟s classification performance. The classification accuracy obtained indicates that the
classifier has ability to correctly classify credit applications thereby identifying “bad” credit applications at an early
stage hence reducing loss of revenue. Implementation of such model in Kenyan Commercial banks can be helpful
for the decision making process.
Citation
Masters of science in computer sciencePublisher
University of Nairobi School of Computing and Informatics