chineduok / creditworthiness Goto Github PK

Udacity Nanodegree Project

The Project focuses of predicting default risk of 500 loan applications to a bank.The R scripts for the different phases of the project are in the 'analysis' folder.
The final report for is an R markdown with code chunks integrated in the report and the presentation is in PDF.This report can be found in the reports folder.
The input data for building the model is in the data folder. 

The project involved developing a binary classification predictive model from 4 different algorithms to determine if a customer is creditworthy or non-creditworthy.
It required hyperparameter tuning of the models to get the optimal model for each algorithm and resampling to select the model that has the best performance for our prediction.

The project had initially been developed using Alteryx during the course. The motivation was to see if the results can be replicated in R.