Credit Risk Evaluation in Banking and Lending Sectors Using Neural Network Model

Abstract

This study presents the results of an experiment made with google colaboratory. The Google Colaboratory is also known as google colab it is a free online cloud-based Jupyter notebook environment that allows us to train our machine learning and deep learning models on CPUs, GPUs, and TPUs. We can determine the accuracy of the credit risk evaluation based on the dataset that was run in google colab using a neural network. The dataset has been thoroughly evaluated in creating a test harness for evaluating candidate models by calculating accuracy using k-fold cross-validation. When compared to a single train-test split, the k-fold cross-validation technique provides a reasonable general approximation of model performance. Based on the result, the correct number of rows was loaded, and through the one-hot encoding of the categorical input variables, it increased the number of input variables from 20 to 61. That suggests that the 13 categorical variables were encoded into a total of 54 columns. The result of the evaluation can help the banking and other financial sectors to assess a person before they are given a loan and if they can pay on time. This is a big help to reduce the losses of a company. Not only in the banking and other financial sectors as well as in lenders of goods.