02_Deployment Loan Request

This workflow deploys a classification model (a Random Forest in this case) to approve or disapprove loan requests.

Previously the model has been trained on 80% of the data from the German Credit Data Set, provided by University of California Archive for Machine Learning and Intelligent Systems, to predict whether an applicant is credit-worthy (1) or risky (2). The loans had been issued for credit-worthy applicants and rejected for risky applicants.

Now we reuse that model, to make the same prediction for new loan applications. Today's loan applications are 100 and are contained in file "daily loan requests.csv". Obviously, since they are new there is not target variable already assigned to them. So we use the model to make the prediction. If the applicant is esteemed credit-worthy (1) then the loan is granted, if risky (2) then the loan request is rejected. The workflow implements the following steps:
1. read data "daily loan requests.csv"
2. Import model
3. perform prediction to either credit-worthy or risky

Read more on the topic Credit Scoring on the KNIME Blog: https://www.knime.com/blog/how-to-do-credit-scoring

External resources

• KNIME Blog: Credit Scoring