Credit Risk Prediction

Credit Risk Predictor

Just an incredibly simple responsive website
which contains a statistical application.

The Credit Risk Predictor(CRP) is a logistic regression model based on the data from UC IRVINE Machine Learning Repository. The dataset contains data from 1973 to 1975 with 700 good and 300 bad credits plus 20 predictor variables. It has been provided as part of a collection of datasets from the "STATLOG" project, a program launched by the EU.

Things the App can do

The following is a short list of the web application's capabilities.

Generation of a logistic regression model
Estimation of credit compliance
Evaluation using sensitivity and specificity
Free selectable threshold

Credit Risk Prediction

In this section you can set the parameters and threshold necessary for model prediction.

Account Status:

Credit duration(in months):

Amount(in EUR):

Credit history:
Purpose:
Savings:
Employment duration:
Installment rate:
Other debtors:
Property:
Other installment plans:
Housing:
Telephone:
Foreign worker:

Threshold: %

Sensitivity: %
Specificity: %