Machine Learning Intermediate 9 min

ML Scoring Models: From Raw Data to Credit Decisions

Machine learning methods applied to credit scoring have shifted from theoretical research to standard practice. The models in this program reflect that shift — built from data patterns, validated on real portfolios, and interpreted with tools that regulators and risk teams actually use.

5 weeks duration
14! seats left
4.5 / 5 avg rating (105 reviews)
ML Scoring Models: From Raw Data to Credit Decisions program cover
1,490 CAD
One-time payment, lifetime access to materials and code repositories
enroll now
only 14 spots remaining
format Intermediate
duration 5 weeks
category Machine Learning
reading time 9 min
price 1,490 CAD

about this program

Credit scoring with machine learning sits at the intersection of statistics, software engineering, and financial regulation. Getting one part wrong can mean rejecting creditworthy applicants or underpricing risk at scale.

What this course covers

You will work through the complete model development cycle: sourcing and cleaning bureau data, handling class imbalance, selecting and tuning classifiers, and generating outputs that satisfy internal audit requirements.

The course uses real-world tabular datasets with the kinds of gaps and inconsistencies you encounter in production — not cleaned Kaggle samples.

Technical stack used

All exercises run in Python using scikit-learn, LightGBM, and Optuna for hyperparameter search. You will also use SHAP for explainability, which is increasingly required under fair lending frameworks in Canada and the EU.

Scoring is not just about AUC. A model that cannot be explained to a compliance officer is a model that cannot be deployed.

Who this is designed for

Analysts and data scientists who already work with structured data and want to move into credit risk or improve existing scoring pipelines. Some prior experience with pandas and basic classification models is assumed.

4.5
based on 105 reviews
what you receive
  • certificate of completion
  • model code and datasets
  • access to Q&A sessions
  • peer review exercises

program curriculum

structured by topic period and learning outcome

foundation
Credit data fundamentals
Working with bureau data, application data, and behavioural variables. Data quality checks and missing-value treatment.
Variable selection and encoding
Weight of evidence encoding, information value, coarse classification — the tooling that still dominates production environments.
model development
Logistic regression as baseline
Why lasso-regularised logistic regression remains a strong default. Scorecard conversion and odds calibration.
Gradient boosting for scoring
XGBoost and LightGBM applied to imbalanced credit datasets. Hyperparameter search without data leakage.
Survival models and through-the-door populations
Handling sample bias from reject inference. Survival analysis for time-to-default outcomes.
validation and deployment
Model performance metrics
Gini coefficient, KS statistic, PSI, and the distinction between discrimination and calibration in regulatory contexts.
Explainability with SHAP
Producing audit-ready explanations for decisions made by tree ensembles. SHAP values in practice, not just theory.
Monitoring and drift detection
Tracking population stability post-deployment. Triggers for model rebuild vs recalibration.
  1. Data Foundations

    Week 1

    Credit bureau file formats, application data schemas, handling missing values in financial features, time-based train/test splitting.

  2. Feature Engineering for Risk

    Week 2

    Weight of Evidence encoding, monotonic binning, delinquency trajectory features, vintage analysis.

  3. Model Selection and Calibration

    Week 3

    Logistic regression baselines, gradient boosting, probability calibration, Gini and KS metrics.

  4. Explainability and Adverse Action

    Week 4

    SHAP values in scoring context, generating reason codes, documenting model decisions for regulatory review.

  5. Deployment and Monitoring

    Week 5

    Score card export, PSI-based drift detection, setting up automated retraining triggers.

questions? write to [email protected]

ready to work with REAL scoring data?

The program runs on actual credit portfolio datasets under controlled conditions. You leave with code, documentation, and a validated model — not slides. Enrolment closes when seats fill.

duration 5 weeks
14 seats remaining
Your data preferences

We use data to keep things running and to improve your quiz experience: