RISKY BANK LOANS PREDICTION THROUGH DATA MINING TECHNIQUES - AN EMPIRICAL COMPARATIVE STUDY
Keywords:
Data Mining, Classification, Prediction, Decision Tree, Naïve Bayes classification, Random ForestAbstract
nowadays, we are witnessing the financial crisis in the banking sector due to many factors and one amongst
those most serious factors is the risky loans. Understanding the customer behavior is more crucial in this context. The
volumes of the data being generated due to banking transactions are increasing day by day. To understand this huge
volume of data, we need to adopt the data mining techniques to get the insights of the data and take the proper decisions.
In this paper, we focus on developing a loan approval model which in turn helps in the prediction of the risky loans using
various data mining techniques like Naïve Bayes, Decision Trees and Random Forest with implementation in R language
and their prediction accuracies are compared further to select the best algorithm.
