A New Framework for Loan Defaulter System Using Support Vector Machine Over a Conditional Random Field Algorithm for a Banker

Abstract

The aim of the study is to predict loan defaulters case in the banking system in order to detect loan defaulters using a novel machine learning method. Here we are using some two algorithms. Materials and Methods: Datasets are downloaded from the kaggle website to train using Novel Support Vector algorithms by applying in 346 records having 10 attributes sample size =250. The model is to analyze and predict the loan defaulters using Support Vector Machine and Random Field algorithm. Results and Discussion: The data was split into training and test datasets. The Random Field algorithm has shown accuracyhas (0.58) with (p<0.05) Conclusion: When compared to the Random Field algorithm (RFA) and Novel Support Vector Machine (SVM) has given best accuracy in predicting loan defaulters' cases using the dataset.