Create-a-binary-classification-model-for-a-loan-decision-app-deployed-in-Azure

Aim of the project :

This project consists in :

• Creating a binary classfication model to predict if a client can get a loan or not to buy a house/apartment.
  
• Deploying the classification model in the cloud via an API.
  
• Creating a Web application deployed in the cloud so that it can be easy to get the output of the classification model for a none-technical person by clicking on the Predict box.

Walkthrough of the Web application :

• The input data are fed via a drop down menu by selecting the index of the client :
  

• By clicking on the Explain prediction button, the user can explain to the client the features that explains the decision using a waterfall SHAP plot.
  

• Finally, by clicking on the Compare with other customers button, the user can illustrate to the client where the client position is compared to other clients based on a number of features, such as income, credit amount etc.
  
  

Videoclip of the Web application once deployed in Azure :

Dashboard_video_Github_short_2.mov