Credit Scoring Using Alternate Data

The objective of this internship is to develop a Retail Credit Risk Scoring Model utilizing different alternate data sources and this model can give a credit score for people who has limited credit footprint. The intern will work on checking alternate data sources availability, Data Preparation, Feature selection, Evaluating models, tuning the model and documentations. This research project also involves several machine learning and deep learning techniques, like XG Boost, Random forest and Recurrent Neural networks. This project will help ICICI Bank in expanding its credit penetration, especially through Digital channels. It can also provide more timely information and improve assessments of creditworthiness for ICICI bank’s customers.

Faculty Supervisor:

Andrei Badescu;Sheldon Lin

Student:

Partner:

ICICI Bank Canada

Discipline:

Computer science

Sector:

Finance and Insurance

University:

University of Toronto

Program: