Worked notebooks

The repository includes self-contained notebook folders under notebooks/. Each folder contains an .ipynb notebook and, where available, matching .py, .html, data, and metadata files.

Folder

Focus

00_quickstart

End-to-end fitting, pattern extraction, and explanation walkthrough.

01_benchmark_baselines

Comparison against common tabular baselines; refreshed in v1.1.7 with the current baseline run and HUGIML grid settings.

02_hug_vs_ebm

HUGIML profiles versus EBM-style additive shape functions.

03_modeling_special_cases

Multiclass targets, imbalance, high-cardinality categoricals, adaptive binning, and pruning.

04_credit_risk

Credit-risk governance using German Credit-style data.

05_aml

Anti-money-laundering suspicious transaction patterns.

06_mobile_money

Mobile-money fraud review signals.

07_basel_ca

Basel capital-adequacy style risk analytics.

08_clinical

Clinical classification and healthcare pattern explanations.

09_insurance

Insurance underwriting risk-selection patterns.

10_medicare

Medicare program-integrity review.

11_workforce_analytics

Workforce attrition and HR risk patterns.

Publishing notebooks

For Read the Docs, keep notebooks as repository examples and link to their HTML exports rather than executing all notebooks during documentation builds. This keeps documentation builds fast and avoids optional data or long-running benchmark dependencies.