Prophecy of loan approval by comparing Decision Tree with Logistic Regression, Random Forest, KNN for better Accuracy.

Authors

  • B. Aditya
  • V. Nagaraju

DOI:

https://doi.org/10.47750/pnr.2022.13.S04.87

Keywords:

Classification, Novel Penalty Based Approach, Logistic Regression, Decision Tree, K-Nearest Neighbor, Random Forest, Loan Approval, Machine Learning, Data Mining.

Abstract

Aim: The aim of the work is to evaluate the accuracy and cross validation in predicting loan approval using Novel Penalty Based Approach Logistic Regression (LR), Random Forest (RF), Decision Tree (DT) and K-Nearest Neighbor (KNN) Classification algorithms. Materials and Methods: The classification algorithm is invoked on a loan approval dataset consisting of 615 records. A framework for forecasting loan approval in the banking sector has been proposed and developed that compares Novel Penalty Based Approach Logistic Regression, decision tree, Random Forest and K-Nearest Neighbor classifiers. Sample size was calculated as 55 in each group using G powers. Sample size was calculated using clinical analysis, with alpha and beta values of 0, 05 and 0. 5, 95% confidence, 80% pre-test g power and enrolment ratio 1.

Results: The Novel Penalty Based Approach Logistic Regression classifier produces 77.27% accuracy in predicting the Loan Approval on the data set, whereas the Random Forest, K-Nearest Neighbor, Decision Tree classifiers produce 76.62%, 72.27%, 72.07% respectively. The significant value is 0.016. Hence Novel Penalty Based Approach Logistic Regression is better than KNN, RF, DT classifiers. Conclusion: The results show that the performance of Novel Penalty Based Approach Logistic Regression is better when compared with KNN, RF and DT in terms of both cross validation and accuracy.

Downloads

Published

2022-09-27

Issue

Section

Articles

How to Cite

Prophecy of loan approval by comparing Decision Tree with Logistic Regression, Random Forest, KNN for better Accuracy. (2022). Journal of Pharmaceutical Negative Results, 759-768. https://doi.org/10.47750/pnr.2022.13.S04.87