Skip to content

Initial Release – H3CCost Framework for Institutional Risk Modeling

Latest

Choose a tag to compare

@motuekely-byte motuekely-byte released this 12 Sep 08:40
· 2 commits to main since this release
cf1a6c7

This release includes the first complete implementation of the H3CCost framework, a cost-sensitive classification system tailored for institutional risk in banking. It integrates strategic cost logic into supervised learning models, including:

  • XGBoost, SVM, Logistic Regression, Naive Bayes, and MLP variants
  • Cost breakdown analysis across multiple datasets
  • Residual institutional loss mapping
  • Comparative performance metrics (Accuracy, ROC AUC, Cost Efficiency)

The codebase supports reproducibility and extension for regulated domains such as finance, credit scoring, and compliance-sensitive environments.

This version aligns with the manuscript submitted to Taylor & Francis and is intended for archival and citation purposes.

DOI