Machine-Learning

The purpose of this Supervised (Classification) learning algorithms is to estimate the prospective borrower's probability of default for their loans which utilized past dataset to make future predictions as accurate as possible

Since lending is one of the most important activities in the financial industry such as banking and finance (Retail banking, Commerical banking or Corporate banking...etc)

The project is based on a loan dataset from "LendingClub bank", a large dataset from 2017-2018 that contains over 100k observations each with 150+ variables

things we are trying to predict: 1. Creditworthiness of the borrower 2. How likely will he or she default our loan

We evaluated different machine learning models such as (KNN, Naïve Bayes, Decision Tree & SVM) and decided to select KNN as the final model based on five important ratios (Recall, Accuracy, Precision, F1-score, ROC)