This lesson explores the application of data analytics and statistical modeling to credit risk assessment and portfolio management.

3.1 The Data Analytics Process in Credit

The University of Southampton’s Credit Risk & Data Analytics module defines a structured three-step process for credit analytics :

  1. Data Pre-processing: Sampling, handling missing values, and outlier treatment .

  2. Data Mining: Applying analytical techniques to identify patterns .

  3. Post-processing: Interpreting and presenting results .

3.2 Types of Data Analytics

The module distinguishes three types of analytics for credit applications :

  • Descriptive/Diagnostic Analytics: Understanding what happened and why (e.g., clustering, association rules) .

  • Predictive Analytics: Forecasting future outcomes (e.g., regression and classification) .

  • Prescriptive Analytics: Recommending actions (e.g., optimization models) .

3.3 Credit Scoring Models

Credit scoring is a primary application of data analytics in credit management . The University of Southampton module emphasizes :

  • Classification Approaches: Logistic regression, decision trees for scorecard development .

  • Performance Measurement: ROC curves, Lift, Gini coefficients .

  • Software Applications: Students work with “world-class software” to develop predictive scorecards .

Bharathidasan University’s FinTech curriculum further highlights “Credit Scoring Innovations” as a key topic, including the role of AI and alternative data in enhancing credit assessment .

PreviousNext