Loan classification using random forest algorithm and comparative analysis with other classifiers

The study aims to analyze the application of the random forest algorithm in addressing the loan classification issue. Furthermore, it intends to perform a comparative analysis by juxtaposing the outcomes with those derived from logistic regression, feedforward neural network, and deep feedforward neural network models. The research determined the ideal maximum number of input indicators and the ideal number of trees in the ensemble when utilizing the random forest algorithm. Additionally, it explored the impact of alternative data partitioning into training and test sets on the accuracy of model forecasting with the random forest algorithm. In conclusion, a strategy for addressing the loan classification issue using the classifiers studied has been proposed.